1755 lines
270 KiB
Plaintext
1755 lines
270 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "dfb11bac",
|
|
"metadata": {
|
|
"id": "dfb11bac"
|
|
},
|
|
"source": [
|
|
"\n",
|
|
"<div class=\"alert alert-block alert-info\">\n",
|
|
"<h1> Widzenie komputerowe </h1>\n",
|
|
"<h2> 01. <i>Wprowadzenie do widzenia komputerowego</i> [laboratoria]</h2> \n",
|
|
"<h3>Andrzej Wójtowicz (2021)</h3>\n",
|
|
"</div>\n",
|
|
"\n",
|
|
""
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "7c64b51d",
|
|
"metadata": {
|
|
"id": "7c64b51d"
|
|
},
|
|
"source": [
|
|
"# Biblioteka OpenCV\n",
|
|
"\n",
|
|
"Podczas zajęć będziemy poruszali zagadnienia związane z [widzeniem komputerowym](https://en.wikipedia.org/wiki/Computer_vision) (ang. *computer vision*, CV). Tę tematykę możemy traktować jako rozwinięcie, czy też bardziej zaawansowaną formę [prztwarzania obrazów](https://en.wikipedia.org/wiki/Digital_image_processing) (ang. *image processing*, IP), gdzie tym razem będziemy starali się wyciągnąć pewną bardziej zaawansowaną wiedzę płynącą z obrazów statycznych lub wideo (cf. dyskusja na [Artificial Intelligence Stack Exchange](https://ai.stackexchange.com/a/13588)). Przedmiot ma formę laboratoryjną, zatem główną dyskusję dotyczącą zakresu obu dziedzin zostawimy w ramach dodatkowej literatury uzupełniającej.\n",
|
|
"\n",
|
|
"Standardem dla algorytmów z dziedzin IP/CV jest biblioteka [OpenCV](https://opencv.org/), która implementuje wiele z tych algorytmów oraz jest aktywnie rozwijana przez społeczność. Sama biblioteka posiada interfejsy do wielu języków programowania, natomiast my skupimy się na języku [Python](https://www.python.org/), który będzie dla nas idealny na potrzeby intensywnego prototypowania. Dokumentację online będzie głównie prowadziła do języka C++, ponieważ nie ma dedykowanej online dla Pythona, ale argumenty funkcji i metod są analogiczne.\n",
|
|
"\n",
|
|
"Początkowe zajęcia będą głównie dotyczyły zagadnień IP, tak aby zapoznać się z biblioteką OpenCV, a dalsze zajęcia będą już związane z CV.\n",
|
|
"\n",
|
|
"## Instalacja\n",
|
|
"\n",
|
|
"Materiały do zajęć Jupyter Notebook są tworzone na serwerze JupyterHub z kernelem Python 3. Pominiemy tutaj tworzenie wirtualnego środowiska, jednak należy mieć na uwadze, że poniższe polecenia mogą być też przydatne podczas próby uruchomienia notebooków lub programów na własnym komputerze.\n",
|
|
"\n",
|
|
"Poniższe polecenie wyświetla używaną wersję Pythona:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 1,
|
|
"id": "8fda1098",
|
|
"metadata": {
|
|
"colab": {
|
|
"base_uri": "https://localhost:8080/"
|
|
},
|
|
"executionInfo": {
|
|
"elapsed": 756,
|
|
"status": "ok",
|
|
"timestamp": 1666007608272,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "8fda1098",
|
|
"outputId": "2549a835-4457-401c-c5ce-1d2d4f0186b2"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"3.7.14\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"import platform\n",
|
|
"print(platform.python_version())"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "c8e7b1e3",
|
|
"metadata": {
|
|
"id": "c8e7b1e3"
|
|
},
|
|
"source": [
|
|
"Niezbędne moduły zainstalujemy poprzez menadżer `pip`. Sama biblioteka OpenCV, abstrahując od np. paczek debianowych, posiada [4 możliwe opcje instalacji](https://pypi.org/project/opencv-contrib-python/). My zainstalujemy pełną wersję tej biblioteki, a dodatkowo doinstalujemy pakiety związane m.in. wyświetlaniem grafiki oraz uczeniem maszynowym."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"id": "78e42f58",
|
|
"metadata": {
|
|
"colab": {
|
|
"base_uri": "https://localhost:8080/"
|
|
},
|
|
"executionInfo": {
|
|
"elapsed": 14652,
|
|
"status": "ok",
|
|
"timestamp": 1666007622922,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "78e42f58",
|
|
"outputId": "b533f0c4-44f7-4792-ff81-d94ec30e483c"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
|
|
"Collecting opencv-contrib-python==4.5.3.56\n",
|
|
" Downloading opencv_contrib_python-4.5.3.56-cp37-cp37m-manylinux2014_x86_64.whl (56.1 MB)\n",
|
|
"\u001b[K |████████████████████████████████| 56.1 MB 1.2 MB/s \n",
|
|
"\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (1.21.6)\n",
|
|
"Requirement already satisfied: scipy in /usr/local/lib/python3.7/dist-packages (1.7.3)\n",
|
|
"Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (3.2.2)\n",
|
|
"Requirement already satisfied: scikit-learn in /usr/local/lib/python3.7/dist-packages (1.0.2)\n",
|
|
"Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (1.4.4)\n",
|
|
"Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (0.11.0)\n",
|
|
"Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (2.8.2)\n",
|
|
"Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib) (3.0.9)\n",
|
|
"Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from kiwisolver>=1.0.1->matplotlib) (4.1.1)\n",
|
|
"Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.7/dist-packages (from python-dateutil>=2.1->matplotlib) (1.15.0)\n",
|
|
"Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from scikit-learn) (3.1.0)\n",
|
|
"Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit-learn) (1.2.0)\n",
|
|
"Installing collected packages: opencv-contrib-python\n",
|
|
"Successfully installed opencv-contrib-python-4.5.3.56\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"!pip3 install --user --disable-pip-version-check opencv-contrib-python==4.5.3.56 numpy scipy matplotlib scikit-learn"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "1d241431",
|
|
"metadata": {
|
|
"id": "1d241431"
|
|
},
|
|
"source": [
|
|
"Sama biblioteka posiada też własne moduły związane z wyświetlaniem grafiki (moduł *HighGUI*) oraz z uczeniem maszynowym (moduł *ml*), jednak my raczej nie będziemy z nich korzystać podczas zajęć (aczkolwiek mogą być przydatne podczas realizacji projektu).\n",
|
|
"\n",
|
|
"Wszystkie moduły, algorytmy i zmienne są dostępne z poziomu modułu `cv2`:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 3,
|
|
"id": "8eca2db5",
|
|
"metadata": {
|
|
"executionInfo": {
|
|
"elapsed": 406,
|
|
"status": "ok",
|
|
"timestamp": 1666007623321,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "8eca2db5"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"import cv2 as cv"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "82701dba",
|
|
"metadata": {
|
|
"id": "82701dba"
|
|
},
|
|
"source": [
|
|
"Poniższym poleceniem możemy również sprawdzić wersję zainstalowanej biblioteki:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 4,
|
|
"id": "f34f551a",
|
|
"metadata": {
|
|
"colab": {
|
|
"base_uri": "https://localhost:8080/"
|
|
},
|
|
"executionInfo": {
|
|
"elapsed": 7,
|
|
"status": "ok",
|
|
"timestamp": 1666007623322,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "f34f551a",
|
|
"outputId": "1aa86a57-7041-4e80-9be2-3543df45bb7f"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"4.6.0\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"print(cv.__version__)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "7a39875c",
|
|
"metadata": {
|
|
"id": "7a39875c"
|
|
},
|
|
"source": [
|
|
"## Obrazy jako tablice\n",
|
|
"\n",
|
|
"Obraz możemy traktować jako standardową tablicę [NumPy](https://www.numpy.org/) zawierającą dane dotyczące danych pikseli. Im większa liczba pikseli w obrazie, tym większa jest jego rozdzielczość. Intuicyjnie można przyjąć, że piksele są niewielkimi blokami informacji ułożonymi w postaci siatki dwuwymiarowej, a głębokość piksela odnosi się do informacji o kolorze. Aby obraz mógł być przetworzony przez komputer, to taki obraz musi zostać przekonwertowany na postać binarną. Kolor obrazu można obliczyć w następujący sposób:\n",
|
|
"\n",
|
|
"Liczba kolorów (odcieni) = 2<sup>*bpp*</sup>, gdzie *bpp* oznacza bity na piksel.\n",
|
|
"\n",
|
|
"Większa liczba bitów na piksel daje więcej możliwych kolorów na obrazach.\n",
|
|
"\n",
|
|
"| Bity na piksel | Liczba kolorów |\n",
|
|
"| :------------: | :------------: |\n",
|
|
"| 1 | 2<sup>1</sup> = 2 |\n",
|
|
"| 2 | 2<sup>2</sup> = 4 |\n",
|
|
"| 3 | 2<sup>3</sup> = 8 |\n",
|
|
"| 4 | 2<sup>4</sup> = 16 |\n",
|
|
"| 8 | 2<sup>8</sup> = 256 |\n",
|
|
"| 16 | 2<sup>16</sup> = 65536 |\n",
|
|
"\n",
|
|
"Spójrzmy na trzy typowe reprezentacje obrazów.\n",
|
|
"\n",
|
|
"### Obraz czarno-biały\n",
|
|
"\n",
|
|
"Obraz binarny składa się z 1 bita na piksel, a więc może mieć tylko dwa możliwe kolory, tj. czarny lub biały. Kolor czarny jest reprezentowany przez wartość 0, a 1 oznacza biel (czasem w użyciu są reprezentacje, które mają odwrotne wartości czerni i bieli).\n",
|
|
"\n",
|
|
"\n",
|
|
"\n",
|
|
"### Obraz w skali odcieni szarości\n",
|
|
"\n",
|
|
"Obraz w skali szarości składa się z 8 bitów na piksel. Oznacza to, że może mieć 256 różnych odcieni, przy czym 0 pikseli będzie reprezentować kolor czarny, a 255 oznacza biel.\n",
|
|
"\n",
|
|
"\n",
|
|
"\n",
|
|
"### Obraz kolorowy\n",
|
|
"\n",
|
|
"Kolorowe obrazy w standardowej formie są reprezentowane jako połączenie barwy czerwonej, niebieskiej i zielonej - wszystkie inne kolory można uzyskać, mieszając te podstawowe kolory we właściwych proporcjach.\n",
|
|
"\n",
|
|
"\n",
|
|
"\n",
|
|
"Kolorowy obraz składa się również z 8 bitów na piksel, z tym że na taki obraz składają się 3 kanały (czerwony, zielony i niebieski). W rezultacie 256 różnych natężeń danego koloru podstawowego można przedstawić za pomocą 0 oznaczającego najmniej intensywny i 255 najbardziej intensywny. Dla poniższego obrazu pawiana:\n",
|
|
"\n",
|
|
"\n",
|
|
"\n",
|
|
"podstawowe parametry dotyczące wymiarów prezentują się następująco:\n",
|
|
"\n",
|
|
"```plaintext\n",
|
|
"Shape\n",
|
|
"(288, 289, 3)\n",
|
|
"288: Pixel height (wysokość w pikselach)\n",
|
|
"289: Pixel width (szerokość w pikselach)\n",
|
|
"3: color channel (liczba kanałów)\n",
|
|
"```\n",
|
|
"\n",
|
|
"Taki obraz możemy reprezentować w postaci trójwymiarowej tablicy:\n",
|
|
"\n",
|
|
""
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "650cae81",
|
|
"metadata": {
|
|
"id": "650cae81"
|
|
},
|
|
"source": [
|
|
"## Wczytywanie obrazów\n",
|
|
"\n",
|
|
"Przy użyciu funkcji [`cv.imread()`](https://docs.opencv.org/4.5.3/d4/da8/group__imgcodecs.html#ga288b8b3da0892bd651fce07b3bbd3a56) możemy odczytać obraz. Pierwszy argument to lokalizacja pliku. Obraz powinien znajdować się w katalogu roboczym lub powinna zostać podana ścieżka bezwzględna do pliku.\n",
|
|
"\n",
|
|
"Drugi argument (opcjonalny) jest flagą oznaczającą w jaki sposób obraz powinien zostać wczytany:\n",
|
|
"\n",
|
|
"* [`cv.IMREAD_COLOR`](https://docs.opencv.org/4.5.3/d4/da8/group__imgcodecs.html#gga61d9b0126a3e57d9277ac48327799c80af660544735200cbe942eea09232eb822) - wczytuje kolorowy obraz z pominięciem przezroczystości; flaga domyślna,\n",
|
|
"* [`cv.IMREAD_GRAYSCALE`](https://docs.opencv.org/4.5.3/d4/da8/group__imgcodecs.html#gga61d9b0126a3e57d9277ac48327799c80ae29981cfc153d3b0cef5c0daeedd2125) : wczytuje obraz w skali odcieni szarości,\n",
|
|
"* [`cv.IMREAD_UNCHANGED`](https://docs.opencv.org/4.5.3/d4/da8/group__imgcodecs.html#gga61d9b0126a3e57d9277ac48327799c80aeddd67043ed0df14f9d9a4e66d2b0708) : wczytuje obraz razem z kanałem alfa (przezroczystość).\n",
|
|
"\n",
|
|
"Zamiast trzech powyższych flag można alternatywnie przekazać odpowiednio 1, 0 lub -1."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 17,
|
|
"id": "d7fccc7d",
|
|
"metadata": {
|
|
"executionInfo": {
|
|
"elapsed": 616,
|
|
"status": "ok",
|
|
"timestamp": 1666007894727,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "d7fccc7d"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image = cv.imread(\"img/baboon.png\", cv.IMREAD_COLOR)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "d9024a2d",
|
|
"metadata": {
|
|
"id": "d9024a2d"
|
|
},
|
|
"source": [
|
|
"Sprawdźmy typ zmiennej:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "78c727fe",
|
|
"metadata": {
|
|
"id": "78c727fe"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"print(type(image))"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "7c488a2d",
|
|
"metadata": {
|
|
"id": "7c488a2d"
|
|
},
|
|
"source": [
|
|
"Sprawdźmy kształt tablicy:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "67bafe94",
|
|
"metadata": {
|
|
"id": "67bafe94"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"print(image.shape)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "68d3db9e",
|
|
"metadata": {
|
|
"id": "68d3db9e"
|
|
},
|
|
"source": [
|
|
"[Typ danych](https://numpy.org/doc/stable/user/basics.types.html) tablicy:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "8463a5df",
|
|
"metadata": {
|
|
"id": "8463a5df"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"print(image.dtype)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "243e1fd8",
|
|
"metadata": {
|
|
"id": "243e1fd8"
|
|
},
|
|
"source": [
|
|
"Jak widać, poszczególne piksele na kanałach są 8-bitowymi liczbami całkowitymi bez znaku."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "185a63ad",
|
|
"metadata": {
|
|
"id": "185a63ad"
|
|
},
|
|
"source": [
|
|
"## Wyświetlanie obrazów przy pomocy Matplotlib\n",
|
|
"\n",
|
|
"W notebooku Jupyter najwygodniej wyświetla się obrazy przy pomocy biblioteki [Matplotlib](https://matplotlib.org/), a dokładniej modułu `pyplot`:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "c4516873",
|
|
"metadata": {
|
|
"id": "c4516873"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"import matplotlib.pyplot as plt\n",
|
|
"import matplotlib\n",
|
|
"%matplotlib inline"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "7e320379",
|
|
"metadata": {
|
|
"id": "7e320379"
|
|
},
|
|
"source": [
|
|
"Przed wyświetleniem obrazu musimy dokonać drobnej konwersji obrazu. OpenCV domyślnie wczytuje obraz w formacie BGR, natomiast Matplotlib pracuje na obrazie w formacie RGB. Do konwersji użyjemy funkcji [`cv.cvtColor()`](https://docs.opencv.org/4.5.3/d8/d01/group__imgproc__color__conversions.html#ga397ae87e1288a81d2363b61574eb8cab), której pierwszym argumentem jest konwertowany obraz, a drugim sposób konwersji (w tym wypadku definiowany przez stałą [`cv.COLOR_BGR2RGB`](https://docs.opencv.org/4.5.3/d8/d01/group__imgproc__color__conversions.html#gga4e0972be5de079fed4e3a10e24ef5ef0ad3db9ff253b87d02efe4887b2f5d77ee)). *Nota bene*, czasami lepiej jest operować w przestrzeni barw [HSV](https://pl.wikipedia.org/wiki/HSV_(grafika)) niż RGB."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "bd083db2",
|
|
"metadata": {
|
|
"id": "bd083db2"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image2 = cv.cvtColor(image, cv.COLOR_BGR2RGB)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "f8b4ae77",
|
|
"metadata": {
|
|
"id": "f8b4ae77"
|
|
},
|
|
"source": [
|
|
"Po przekonwertowaniu obrazu możemy wyświetlić go przy pomocy `pyplot`. Użyjemy do tego funkcji [`matplotlib.pyplot.imshow()`](https://matplotlib.org/api/pyplot_api.html#matplotlib.pyplot.imshow):"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "e39ea012",
|
|
"metadata": {
|
|
"id": "e39ea012"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"plt.imshow(image2)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "fc1c34db",
|
|
"metadata": {
|
|
"id": "fc1c34db"
|
|
},
|
|
"source": [
|
|
"Zwróćmy uwagę, że piksel o współrzędnych `(0, 0)` znajduje się w lewym górnym rogu.\n",
|
|
"\n",
|
|
"Można też obejść się bez konwersji i po prostu odwrócić w locie kanały w tablicy `numpy`:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "aa374723",
|
|
"metadata": {
|
|
"id": "aa374723"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"plt.imshow(image[..., ::-1])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "232c3da7",
|
|
"metadata": {
|
|
"id": "232c3da7"
|
|
},
|
|
"source": [
|
|
"Sprawdźmy kolejny plik:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "bccd478c",
|
|
"metadata": {
|
|
"id": "bccd478c"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"x_small = cv.imread(\"img/x_small.jpg\", cv.IMREAD_GRAYSCALE)\n",
|
|
"plt.imshow(x_small, cmap = 'gray');"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "2ba47521",
|
|
"metadata": {
|
|
"id": "2ba47521"
|
|
},
|
|
"source": [
|
|
"Wczytaliśmy obraz w skali odcieni szarości. Podczas wyświetlania obrazu możemy ustawić mapę kolorów (parametr `cmap`). Dodatkowo, jeśli na końcu polecenia damy średnik (`;`), to w wynikowej komórce notebooka nie będzie się wyświetlał zwracany typ.\n",
|
|
"\n",
|
|
"Czasami warto również wyświetlić pasek z informacją o tym jaka wartość odpowiada za dany odcień:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "066b4778",
|
|
"metadata": {
|
|
"id": "066b4778"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"plt.imshow(x_small, cmap = 'gray');\n",
|
|
"plt.colorbar();"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "c561f1ea",
|
|
"metadata": {
|
|
"id": "c561f1ea"
|
|
},
|
|
"source": [
|
|
"## Proste operacje\n",
|
|
"\n",
|
|
"Możemy sprawdzić, że obraz jest w istocie tablicą NumPy o wartościach w zakresie od 0 do 255:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "fd384995",
|
|
"metadata": {
|
|
"id": "fd384995"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"print(x_small)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "794d3cb6",
|
|
"metadata": {
|
|
"id": "794d3cb6"
|
|
},
|
|
"source": [
|
|
"Wracając do samej biblioteki Matplotlib, możemy również np. sterować rozmiarem wyświetlanego obrazu oraz nadać mu tytuł:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "5998b09d",
|
|
"metadata": {
|
|
"id": "5998b09d"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"plt.figure(figsize = [6, 6])\n",
|
|
"plt.imshow(x_small, cmap = 'gray')\n",
|
|
"plt.title('Small X', fontsize = 20);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "059710ee",
|
|
"metadata": {
|
|
"id": "059710ee"
|
|
},
|
|
"source": [
|
|
"Niektóre parametry Matlplotlib (np. `figsize`) możemy ustawić na domyślne przy pomocy [`matplotlib.rcParams`](https://matplotlib.org/stable/tutorials/introductory/customizing.html#matplotlib-rcparams).\n",
|
|
"\n",
|
|
"Podczas wczytywania obrazów zawierających przezroczystość należy zwrócić uwagę na sposób wczytywania obrazu:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "77861dde",
|
|
"metadata": {
|
|
"id": "77861dde"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"img_lfl = cv.imread(\"img/linux_foundation.png\", cv.IMREAD_COLOR)\n",
|
|
"plt.imshow(img_lfl);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "46deb78b",
|
|
"metadata": {
|
|
"id": "46deb78b"
|
|
},
|
|
"source": [
|
|
"Mamy tutaj obraz 4-kanałowy, przez co musimy go wczytać jako `cv.IMREAD_UNCHANGED`, no i dodatkowo odpowiednio go przekonwertować:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "ce0a9061",
|
|
"metadata": {
|
|
"id": "ce0a9061"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"img_lfl = cv.imread(\"img/linux_foundation.png\", cv.IMREAD_UNCHANGED)\n",
|
|
"plt.imshow(cv.cvtColor(img_lfl, cv.COLOR_BGRA2RGBA));"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "9f0f1db3",
|
|
"metadata": {
|
|
"id": "9f0f1db3"
|
|
},
|
|
"source": [
|
|
"Możemy również podzielić obraz na poszczególne kanały i ew. ponownie je połączyć, odpowiednio przy pomocy funkcji [`cv.split()`](https://docs.opencv.org/4.5.3/d2/de8/group__core__array.html#ga0547c7fed86152d7e9d0096029c8518a) i [`cv.merge()`](https://docs.opencv.org/4.5.3/d2/de8/group__core__array.html#ga7d7b4d6c6ee504b30a20b1680029c7b4):"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "60ab993b",
|
|
"metadata": {
|
|
"id": "60ab993b"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"b, g, r = cv.split(image)\n",
|
|
"plt.imshow(b, cmap = 'gray')\n",
|
|
"plt.title('Blue channel', fontsize = 20);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "f8418716",
|
|
"metadata": {
|
|
"id": "f8418716"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"restored_image = cv.merge((r,g,b))\n",
|
|
"plt.imshow(restored_image);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "9ba6f17e",
|
|
"metadata": {
|
|
"id": "9ba6f17e"
|
|
},
|
|
"source": [
|
|
"Dostęp do poszczególnych pikseli odbywa się tak jak w macierzy, czyli podając wiersz i kolumnę. Poniżej możemy zobaczyć zmianę piksela na współrzędnych `(1, 1)`:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "7a018550",
|
|
"metadata": {
|
|
"id": "7a018550"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"plt.imshow(x_small, cmap = 'gray');"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "3aebbb19",
|
|
"metadata": {
|
|
"id": "3aebbb19"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"x_small[1, 1] = 255\n",
|
|
"plt.imshow(x_small, cmap = 'gray');"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "14278fd3",
|
|
"metadata": {
|
|
"id": "14278fd3"
|
|
},
|
|
"source": [
|
|
"Przycinanie odbywa się przez znany w Pythonie tzw. *slicing*, czyli wycinki:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "52b49c59",
|
|
"metadata": {
|
|
"id": "52b49c59"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"plt.imshow(x_small[3:8, 2:], cmap = 'gray');"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "e191d0e8",
|
|
"metadata": {
|
|
"id": "e191d0e8"
|
|
},
|
|
"source": [
|
|
"Możemy również ustawić kolor dla kilku kanałów jednocześnie dla zadanego regionu. Wykonamy to na kopii obrazu:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "94db5076",
|
|
"metadata": {
|
|
"id": "94db5076"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image_edited = image.copy()\n",
|
|
"image_edited[100:200, 150:250] = (128, 0, 128)\n",
|
|
"plt.imshow(image_edited[..., ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "f340e7ec",
|
|
"metadata": {
|
|
"id": "f340e7ec"
|
|
},
|
|
"source": [
|
|
"Przeskalowanie obrazu odbywa się przez [`cv.resize()`](https://docs.opencv.org/4.5.3/da/d54/group__imgproc__transform.html#ga47a974309e9102f5f08231edc7e7529d), w której albo podajemy dokładne docelowe wymiary, albo podajemy współczynniki skalowania na osiach *x* i *y*; możemy też uwzględnić odpowiednią metodę interpolacji. Matplotlib nie wyświetli nam wprost powiększonych obrazków, ale będziemy mogli zauważyć zmianę poprzez np. poprzez zmianę zakresów skali osi *x* i *y*:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "fda0546a",
|
|
"metadata": {
|
|
"id": "fda0546a"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"x_small_resized_1 = cv.resize(x_small, (40, 40))\n",
|
|
"plt.imshow(x_small_resized_1, cmap='gray');"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "e85d632e",
|
|
"metadata": {
|
|
"id": "e85d632e"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"x_small_resized_2 = cv.resize(x_small, None, fx=3, fy=4, interpolation=cv.INTER_CUBIC)\n",
|
|
"plt.imshow(x_small_resized_2, cmap='gray');"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "09a01036",
|
|
"metadata": {
|
|
"id": "09a01036"
|
|
},
|
|
"source": [
|
|
"Obrót obrazu dokonywany jest przez funkcję [`cv.flip()`](https://docs.opencv.org/4.5.3/d2/de8/group__core__array.html#gaca7be533e3dac7feb70fc60635adf441), w której podajemy według której osi ma nastąpić obrót (`0`: *x*, `1`: *y*, `-1`: obie)."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "fd673085",
|
|
"metadata": {
|
|
"id": "fd673085"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image_flipped = cv.flip(image, 0)\n",
|
|
"plt.imshow(image_flipped[..., ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "e4c0f0d8",
|
|
"metadata": {
|
|
"id": "e4c0f0d8"
|
|
},
|
|
"source": [
|
|
"Może pojawić się potrzeba utworzenia nowych obrazów, więc najczęściej używa się do tego biblioteki NumPy:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "fc86f294",
|
|
"metadata": {
|
|
"id": "fc86f294"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"import numpy as np"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "bc67f439",
|
|
"metadata": {
|
|
"id": "bc67f439"
|
|
},
|
|
"source": [
|
|
"Poniższe wywołania pokazują kilka wariantów. Możemy utworzyć pustą trójwymiarową macierz wypełnioną zerami przy pomocy [`numpy.zeros()`](https://numpy.org/doc/stable/reference/generated/numpy.zeros.html):"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "a8cb1c58",
|
|
"metadata": {
|
|
"id": "a8cb1c58"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image_empty = np.zeros((50, 100, 3), dtype='uint8')\n",
|
|
"plt.imshow(image_empty);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "dde3f2b0",
|
|
"metadata": {
|
|
"id": "dde3f2b0"
|
|
},
|
|
"source": [
|
|
"Jeżeli chcemy ustawić jakąś początkową wartość, to możemy przeskalować tablicę jedynek uzyskaną z [`numpy.ones()`](https://numpy.org/doc/stable/reference/generated/numpy.ones.html):"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "7190f1df",
|
|
"metadata": {
|
|
"id": "7190f1df"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image_empty = 128*np.ones((50, 100, 3), dtype='uint8')\n",
|
|
"plt.imshow(image_empty)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "4f785c86",
|
|
"metadata": {
|
|
"id": "4f785c86"
|
|
},
|
|
"source": [
|
|
"Jeżeli nowy obraz powinien mieć takie same wymiary jak inny obraz, to możemy do tej operacji użyć [`numpy.ones_like()`](https://numpy.org/doc/stable/reference/generated/numpy.ones_like.html):"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "0f99dda7",
|
|
"metadata": {
|
|
"id": "0f99dda7"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"empty_from_other = 255*np.ones_like(image)\n",
|
|
"plt.imshow(empty_from_other);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "ac02a038",
|
|
"metadata": {
|
|
"id": "ac02a038"
|
|
},
|
|
"source": [
|
|
"Czasami może być przydatne utworzenie maski zawierającej informację o tym czy np. dany piksel zawiera wartość w danym zakresie. Przy pomocy [`cv.inRange()`](https://docs.opencv.org/4.5.3/d2/de8/group__core__array.html#ga48af0ab51e36436c5d04340e036ce981) możemy sprawdzić zakresy dla trzech kanałów; przy okazji możemy zobaczyć jak zgrupować kilka obrazów jednocześnie (por. [`matplotlib.pyplot.subplot()`](https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.subplot.html)):"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "d536f6e6",
|
|
"metadata": {
|
|
"id": "d536f6e6"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"mask = cv.inRange(image, (0,100,150), (100,150,200))\n",
|
|
"\n",
|
|
"plt.figure(figsize=[10,10])\n",
|
|
"plt.subplot(121)\n",
|
|
"plt.imshow(image[..., ::-1])\n",
|
|
"plt.title(\"Baboon\")\n",
|
|
"plt.subplot(122)\n",
|
|
"plt.imshow(mask, cmap='gray')\n",
|
|
"plt.title(\"Mask\");"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "622a0f0d",
|
|
"metadata": {
|
|
"id": "622a0f0d"
|
|
},
|
|
"source": [
|
|
"Standardowo operujemy w zakresie liczb całkowitych `[0; 255]`, przez co jeśli w wyniku jakiejś operacji mielibyśmy wykroczyć poza zakres, to skończy się to albo operacją modulo albo np. uzyskaniem wartości ujemnych, co ostatecznie przełoży się na błędny wynik.\n",
|
|
"\n",
|
|
"Załóżmy, że chcemy podbić kontrast o 50%. Przemnożenie obrazu o wartość 1.5 zwróci nam tablicę o wartościach liczb rzeczywistych:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "d48cf055",
|
|
"metadata": {
|
|
"id": "d48cf055"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"new_image = image * 1.5\n",
|
|
"print(new_image.dtype)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "b2dbeacc",
|
|
"metadata": {
|
|
"id": "b2dbeacc"
|
|
},
|
|
"source": [
|
|
"W takim wypadku możemy naiwnie spróbować przekonwertować to ponownie do liczb całkowitych, ale wartości powyżej 255 zostaną obliczone według operacji modulo, przez co wynik będzie niezadowalający:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "411dabb9",
|
|
"metadata": {
|
|
"id": "411dabb9"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"new_image = np.uint8(new_image)\n",
|
|
"plt.imshow(new_image[..., ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "1fea50c7",
|
|
"metadata": {
|
|
"id": "1fea50c7"
|
|
},
|
|
"source": [
|
|
"Rozwiązaniem jest przycięcie wartości przy użyciu [`numpy.clip()`](https://numpy.org/doc/stable/reference/generated/numpy.clip.html) do zakresu `[0; 255]` jeśli operujemy na liczbach całkowitych, z ew. znormalizowaniem wartości do zakresu `[0; 1]` jeśli chcemy pracować na liczbach rzeczywistych:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "eac38489",
|
|
"metadata": {
|
|
"id": "eac38489"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"new_image = image * 1.5\n",
|
|
"clipped_new_image = np.clip(new_image, 0, 255)\n",
|
|
"clipped_new_image = np.uint8(clipped_new_image)\n",
|
|
"plt.imshow(clipped_new_image[..., ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "bc2cb3df",
|
|
"metadata": {
|
|
"id": "bc2cb3df"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"new_image = (image * 1.5)/255\n",
|
|
"clipped_new_image = np.clip(new_image, 0, 1)\n",
|
|
"plt.imshow(clipped_new_image[..., ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "a2f8c4a5",
|
|
"metadata": {
|
|
"id": "a2f8c4a5"
|
|
},
|
|
"source": [
|
|
"Jeśli zwiększymy jasność i potencjalnie wyjdziemy poza zakres `[0; 255]`, to ponownie wartości zostaną przekręcone:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "ce823b52",
|
|
"metadata": {
|
|
"id": "ce823b52"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"new_image = image + 50\n",
|
|
"plt.imshow(new_image[..., ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "4be595c9",
|
|
"metadata": {
|
|
"id": "4be595c9"
|
|
},
|
|
"source": [
|
|
"Z drugiej strony należy mieć też na uwadze pewne subtelne konwersje, które mogą odbywać się niejawnie. Poniżej mamy operację, która powoduje użycie danych typu `int16`, przez co następuje niejawna konwersja i przycięcie (jest o tym informacja w ostrzeżeniu):"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "c93528a2",
|
|
"metadata": {
|
|
"id": "c93528a2"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"new_image = image + (-50)\n",
|
|
"plt.imshow(new_image[..., ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "6b3a8f53",
|
|
"metadata": {
|
|
"id": "6b3a8f53"
|
|
},
|
|
"source": [
|
|
"Rozwiązaniem powyższych problemów może być w przypadku zmiany jasności użycie funkcji [`cv.add()`](https://docs.opencv.org/4.5.3/d2/de8/group__core__array.html#ga10ac1bfb180e2cfda1701d06c24fdbd6) i [`cv.substract()`](https://docs.opencv.org/4.5.3/d2/de8/group__core__array.html#gaa0f00d98b4b5edeaeb7b8333b2de353b), a w przypadku zmiany kontrastu użycie [`cv.multiply()`](https://docs.opencv.org/4.5.3/d2/de8/group__core__array.html#ga979d898a58d7f61c53003e162e7ad89f), konwersja do typu np. `float64` i powrót do `uint8` z przycięciem wartości."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "fba240f9",
|
|
"metadata": {
|
|
"id": "fba240f9"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"matrix = np.ones(image.shape, dtype='uint8') * 50\n",
|
|
"\n",
|
|
"image_brighter = cv.add(image, matrix)\n",
|
|
"plt.imshow(image_brighter[:, :, ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "79b54cc1",
|
|
"metadata": {
|
|
"id": "79b54cc1"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image_darker = cv.subtract(image, matrix)\n",
|
|
"plt.imshow(image_darker[:, :, ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "2fe3e9d5",
|
|
"metadata": {
|
|
"id": "2fe3e9d5"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"matrix = np.ones(image.shape, dtype = 'float64')\n",
|
|
"\n",
|
|
"image_lower = np.uint8(cv.multiply(np.float64(image), matrix, scale = 0.8))\n",
|
|
"plt.imshow(image_lower[:, :, ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "65da736b",
|
|
"metadata": {
|
|
"id": "65da736b"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"image_higher = np.uint8(np.clip(cv.multiply(np.float64(image), matrix, scale = 1.2) , 0, 255))\n",
|
|
"plt.imshow(image_higher[:, :, ::-1]);"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "54e22dd0",
|
|
"metadata": {
|
|
"id": "54e22dd0"
|
|
},
|
|
"source": [
|
|
"## Zapisywanie obrazów\n",
|
|
"\n",
|
|
"Funkcja [`cv.imwrite()`](https://docs.opencv.org/4.5.3/d4/da8/group__imgcodecs.html#gabbc7ef1aa2edfaa87772f1202d67e0ce) zapisuje obraz do pliku. Pierwszy argument to nazwa pliku, drugi argument to obraz, który chcemy zapisać. Np. poniższe polecenie spowodowałoby zapisanie obrazu w formacie PNG w katalogu roboczym:\n",
|
|
"\n",
|
|
"```python\n",
|
|
"cv.imwrite('pawian.png', image)\n",
|
|
"```\n",
|
|
"\n",
|
|
"## Anotowanie\n",
|
|
"\n",
|
|
"Sprawdźmy w jaki sposób do pustego obrazu dodać kilka obiektów geometrycznych. Użyjemy do tego funkcji [`cv.line()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga7078a9fae8c7e7d13d24dac2520ae4a2), [`cv.circle()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#gaf10604b069374903dbd0f0488cb43670) , [`cv.rectangle()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga07d2f74cadcf8e305e810ce8eed13bc9), [`cv.ellipse()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga28b2267d35786f5f890ca167236cbc69), [`cv.polylines()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#gaa3c25f9fb764b6bef791bf034f6e26f5) i [`cv.putText()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga5126f47f883d730f633d74f07456c576). Funkcje te będą nadpisywały wejściowy obraz.\n",
|
|
"\n",
|
|
"W powyższych funkcjach pojawiają się wspólne argumenty:\n",
|
|
"\n",
|
|
"* `img` - obraz, na którym będziemy umieszczać obiekty,\n",
|
|
"* `color` - kolor obiektu podany w formacie jako krotka (*tuple*), np. `(255, 0, 0)`; w przypadku obrazów w skali odcieni szarości wystarczy podać wartość skalarną,\n",
|
|
"* `thickness` - grubość linii, okręgu, itp.; w przypadku wartości `-1` w przypadku zamkniętych figur jak okrąg, obiekt zostanie wypełniony wewnątrz wskazanym kolorem; domyślna wartość to `1`,\n",
|
|
"* `lineType` : rodzaj linii, np. 8-sąsiedztwo, antyaliasing, itp.; domyślnie jest 8-sąsiedztwo, natomiast wartość [`cv.LINE_AA`](https://docs.opencv.org/4.5.3/d0/de1/group__core.html#ggaf076ef45de481ac96e0ab3dc2c29a777a85fdabe5335c9e6656563dfd7c94fb4f) włącza antyaliasing.\n",
|
|
"\n",
|
|
"Na początku zaimportujemy bibliotekę NumPy i przy pomocy funkcji [`numpy.zeros()`](https://docs.scipy.org/doc/numpy/reference/generated/numpy.zeros.html) utworzymy pusty 3-warstwowy czarny obraz o rozmiarze 512x512."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "efd40bcf",
|
|
"metadata": {
|
|
"id": "efd40bcf"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"import numpy as np\n",
|
|
"\n",
|
|
"img = np.zeros((512,512,3), np.uint8)\n",
|
|
"\n",
|
|
"plt.imshow(img)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "2b23ae8a",
|
|
"metadata": {
|
|
"id": "2b23ae8a"
|
|
},
|
|
"source": [
|
|
"W dalszej części operowali na obrazie w formacie RGB.\n",
|
|
"\n",
|
|
"### Linie\n",
|
|
"\n",
|
|
"Aby narysować linię, musimy podać początkowe i końcowe współrzędne linii. Przy pomocy [`cv.line()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga7078a9fae8c7e7d13d24dac2520ae4a2) narysujemy na obrazie czerwoną linię od lewego górnego rogu do prawego dolnego rogu o grubości 5 pikseli."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "eecb08e6",
|
|
"metadata": {
|
|
"id": "eecb08e6"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"cv.line(img, (0,0), (511,511), (255,0,0), 5)\n",
|
|
"\n",
|
|
"plt.imshow(img)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "bb1a0101",
|
|
"metadata": {
|
|
"id": "bb1a0101"
|
|
},
|
|
"source": [
|
|
"### Prostokąty\n",
|
|
"\n",
|
|
"Aby narysować prostokąt, potrzebujemy lewego górnego rogu i prawego dolnego rogu prostokąta. Tym razem przy pomocy [`cv.rectangle()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga07d2f74cadcf8e305e810ce8eed13bc9) narysujemy zielony prostokąt w prawym górnym rogu obrazu."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "9d3cc9ad",
|
|
"metadata": {
|
|
"id": "9d3cc9ad"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"cv.rectangle(img, (384,0), (510,128), (0,255,0), 3)\n",
|
|
"\n",
|
|
"plt.imshow(img)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "cdcb4b70",
|
|
"metadata": {
|
|
"id": "cdcb4b70"
|
|
},
|
|
"source": [
|
|
"### Okręgi\n",
|
|
"\n",
|
|
"Aby narysować okrąg, potrzebujemy jego współrzędnych środka i promienia. Przy pomocy [`cv.circle()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#gaf10604b069374903dbd0f0488cb43670) narysujemy okrąg wewnątrz prostokąta narysowanego powyżej."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "48207862",
|
|
"metadata": {
|
|
"id": "48207862"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"cv.circle(img, (447,63), 63, (0,0,255), -1)\n",
|
|
"\n",
|
|
"plt.imshow(img)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "60678ca6",
|
|
"metadata": {
|
|
"id": "60678ca6"
|
|
},
|
|
"source": [
|
|
"### Elipsy\n",
|
|
"\n",
|
|
"Aby narysować elipsę, musimy przekazać kilka argumentów. Jednym argumentem jest położenie środka `(x, y)`. Następnym argumentem jest długość dwóch osi (długość głównej osi i mniejsza długość osi). Kolejny parametr to kąt obrotu elipsy w kierunku przeciwnym do ruchu wskazówek zegara. Kolejne dwa argumenty oznaczają początek i koniec łuku elipsy mierzonego zgodnie z ruchem wskazówek zegara od głównej osi. tj. podanie wartości `0` i `360` daje pełną elipsę. Aby uzyskać więcej informacji, sprawdź dokumentację [`cv.ellipse()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga28b2267d35786f5f890ca167236cbc69). Poniższy kod rysuje pół elipsy na środku obrazu."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "4423e4b0",
|
|
"metadata": {
|
|
"id": "4423e4b0"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"cv.ellipse(img, (256,256), (100,50), 0, 0, 180, 255, -1)\n",
|
|
"\n",
|
|
"plt.imshow(img)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "1219997d",
|
|
"metadata": {
|
|
"id": "1219997d"
|
|
},
|
|
"source": [
|
|
"### Poligony\n",
|
|
"\n",
|
|
"Aby narysować wielokąt, najpierw potrzebujemy współrzędnych wierzchołków. Ustawmy te punkty w tablicy o wymiarach `liczba_wierszy x 1 x 2`, gdzie `liczba_wierszy` to liczba wierzchołków i powinna być typu `int32`. Tutaj narysujemy mały wielokąt z czterema wierzchołkami w kolorze jasnoniebieskim."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "60223ef1",
|
|
"metadata": {
|
|
"id": "60223ef1"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"pts = np.array([[10,5], [20,30], [70,20], [50,10]], np.int32)\n",
|
|
"pts = pts.reshape((-1, 1, 2))\n",
|
|
"cv.polylines(img, [pts], True, (0,255,255))\n",
|
|
"\n",
|
|
"plt.imshow(img)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "cdba71d3",
|
|
"metadata": {
|
|
"id": "cdba71d3"
|
|
},
|
|
"source": [
|
|
"Wielokąt może wydawać się przerywany, ale jeśli powiększylibyśmy wynikowy obraz, to byśmy zobaczyli, że jest on zamknięty.\n",
|
|
"\n",
|
|
"Jeśli trzecim argumentem [`cv.polylines()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#gaa3c25f9fb764b6bef791bf034f6e26f5) jest `False`, otrzymamy niedomknięty poligon.\n",
|
|
"\n",
|
|
"Funkcja `cv.polylines()` może być używana do rysowania wielu linii. Wystarczy utworzyć listę wszystkich linii, które chcemy narysować i przekazać je do funkcji. Wszystkie linie zostaną narysowane osobno. Jest to znacznie szybszy sposób na narysowanie grupy linii niż wywołanie osobno [`cv.line()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga7078a9fae8c7e7d13d24dac2520ae4a2) dla każdej linii.\n",
|
|
"\n",
|
|
"### Tekst\n",
|
|
"\n",
|
|
"Aby umieścić tekst na obrazie, musimy określić następujące rzeczy:\n",
|
|
"\n",
|
|
"* dane tekstowe, które chcemy napisać,\n",
|
|
"* współrzędne miejsca, w którym chcemy umieścić tekst (np. lewy dolny róg, w którym zacznie się tekst),\n",
|
|
"* typ czcionki (sprawdź dokumentację [`cv.putText()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga5126f47f883d730f633d74f07456c576) dla obsługiwanych czcionek),\n",
|
|
"* rozmiar czcionki,\n",
|
|
"* pozostałe rzeczy takie jak kolor, grubość, typ linii, itp.; dla uzyskania lepszego wynikowego wyglądu zaleca się `lineType = `[`cv.LINE_AA`](https://docs.opencv.org/4.5.3/d0/de1/group__core.html#ggaf076ef45de481ac96e0ab3dc2c29a777a85fdabe5335c9e6656563dfd7c94fb4f).\n",
|
|
"\n",
|
|
"Na naszym obrazie umieścimy tekst w kolorze białym."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "2de2bcf6",
|
|
"metadata": {
|
|
"id": "2de2bcf6"
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"cv.putText(img, 'OpenCV', (10,500), cv.FONT_HERSHEY_SIMPLEX, 4, (255,255,255), 2, cv.LINE_AA)\n",
|
|
"\n",
|
|
"plt.imshow(img)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "326c28ac",
|
|
"metadata": {
|
|
"id": "326c28ac"
|
|
},
|
|
"source": [
|
|
"Czasami potrzebujemy jednak dokładnych informacji o wynikowym tekście, tak aby np. odpowiednio go umiejscowić lub umieścić pod nim kontrastowe tło. Do uzyskania tych informacji pomocne są funkcje [`cv.getTextSize()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga3d2abfcb995fd2db908c8288199dba82) i [`cv.getFontScaleFromHeight()`](https://docs.opencv.org/4.5.3/d6/d6e/group__imgproc__draw.html#ga442ff925c1a957794a1309e0ed3ba2c3)."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "eec58cd3",
|
|
"metadata": {
|
|
"id": "eec58cd3"
|
|
},
|
|
"source": [
|
|
"## Wyświetlanie obrazów poza Jupyterem\n",
|
|
"\n",
|
|
"### HighGUI\n",
|
|
"\n",
|
|
"W OpenCV obraz standardowo wyświetla się przy pomocy modułu HighGUI. Przy użyciu funkcji [`cv.imshow()`](https://docs.opencv.org/4.5.3/d7/dfc/group__highgui.html#ga453d42fe4cb60e5723281a89973ee563) możemy wyświetlić obraz w oknie. Okno automatycznie dopasowuje się do rozmiaru obrazu.\n",
|
|
"\n",
|
|
"Pierwszy argument to nazwa okna. Drugi argument to nasz obraz. Możemy utworzyć dowolną liczbę okien, ale z różnymi nazwami okien.\n",
|
|
"\n",
|
|
"```python\n",
|
|
"cv.imshow('image', image)\n",
|
|
"cv.waitKey(0)\n",
|
|
"```\n",
|
|
"\n",
|
|
"Pod Windowsem wyświetli nam się wskazany przez nas obraz, natomiast pod Linuxem mechanizm HighGUI przy wyświetlaniu obrazu udostępnia trochę więcej opcji i informacji o wyświetlanym obrazie.\n",
|
|
"\n",
|
|
"\n",
|
|
"\n",
|
|
"Zwróćmy uwagę, że obraz pojawia się dopiero po poleceniu `cv.waitKey(0)` oraz będzie oczekiwał na wciśnięcie klawisza.\n",
|
|
"\n",
|
|
"[`cv.waitKey()`](https://docs.opencv.org/4.5.3/d7/dfc/group__highgui.html#ga5628525ad33f52eab17feebcfba38bd7) jest funkcją związaną z klawiaturą. Jej argumentem jest czas w milisekundach. Funkcja czeka przez określony czas na każde zdarzenie klawiatury. Jeśli w tym czasie zostanie naciśnięty dowolny klawisz, program przejdzie do dalszych operacji. Jeśli jako argument zostanie przekazane 0, to funkcja czeka bez końca na uderzenie klawisza klawiatury. Można również ustawić wykrywanie konkretnych klawiszy, takich jak naciśnięcie klawisza `<A>`, itp. Poza przetwarzaniem zdarzeń klawiatury, ta funkcja przetwarza również wiele innych zdarzeń GUI, więc **musisz** użyć jej przy wyświetleniu obrazu.\n",
|
|
"\n",
|
|
"Klikając na obraz i wciskając `<Enter>`, konsola zwróci informację liczbową o wciśniętym klawiszu, a ponadto obsługa programu wróci z powrotem do konsoli. Wydając poniższe polecenie usuniemy wszystkie okna:\n",
|
|
"\n",
|
|
"```python\n",
|
|
"cv.destroyAllWindows()\n",
|
|
"```\n",
|
|
"\n",
|
|
"[`cv.destroyAllWindows()`](https://docs.opencv.org/4.5.3/d7/dfc/group__highgui.html#ga6b7fc1c1a8960438156912027b38f481) usuwa wszystkie utworzone przez nas okna, a jeśli chcemy usunąć konkretne okno, należy użyć funkcji [`cv.destroyWindow()`](https://docs.opencv.org/4.5.3/d7/dfc/group__highgui.html#ga851ccdd6961022d1d5b4c4f255dbab34), w której jako argument podajemy dokładną nazwę okna.\n",
|
|
"\n",
|
|
"Istnieje specjalny przypadek, w którym możemy utworzyć okno i później załadować do niego obraz. W takim przypadku określamy czy rozmiar okna jest zmienny, czy nie. Robi się to za pomocą funkcji [`cv.namedWindow()`](https://docs.opencv.org/4.5.3/d7/dfc/group__highgui.html#ga5afdf8410934fd099df85c75b2e0888b). Domyślna flaga to [`cv.WINDOW_AUTOSIZE`](https://docs.opencv.org/4.5.3/d7/dfc/group__highgui.html#ggabf7d2c5625bc59ac130287f925557ac3acf621ace7a54954cbac01df27e47228f), ale jeśli określimy flagę jako [`cv.WINDOW_NORMAL`](https://docs.opencv.org/4.5.3/d7/dfc/group__highgui.html#ggabf7d2c5625bc59ac130287f925557ac3a29e45c5af696f73ce5e153601e5ca0f1), to możemy zmienić rozmiar okna. Będzie to pomocne, gdy obraz jest zbyt duży:\n",
|
|
"\n",
|
|
"```python\n",
|
|
"cv.namedWindow('image', cv.WINDOW_NORMAL)\n",
|
|
"cv.imshow('image', image)\n",
|
|
"cv.waitKey(0)\n",
|
|
"cv.destroyAllWindows()\n",
|
|
"```\n",
|
|
"\n",
|
|
"Interfejs HighGUI czasami pomaga w debugowaniu, ale trzeba mieć też świadomość, że nie jest on specjalnie rozbudowany i zasadniczo nie służy do budowania zaawansowanych interfejsów graficznych. HighGUI obsługuje sygnały idące nie tylko z klawiatury, ale również myszy, a ponadto pozawala na umieszczenie np. sliderów. Krótkie przykłady znajdują się w samouczkach opisujących [`cv.setMouseCallback()`](https://docs.opencv.org/4.5.3/db/d5b/tutorial_py_mouse_handling.html) oraz [`cv.getTrackbarPos()` i `cv.createTrackbar()`](https://docs.opencv.org/4.5.3/d9/dc8/tutorial_py_trackbar.html).\n",
|
|
"\n",
|
|
"### Matplotlib\n",
|
|
"\n",
|
|
"Po przekonwertowaniu obrazu do formatu RGB wyświetlamy go przy pomocy `pyplot`. Używamy przy tym funkcji [`matplotlib.pyplot.imshow()`](https://matplotlib.org/api/pyplot_api.html#matplotlib.pyplot.imshow) oraz [`matplotlib.pyplot.show()`](https://matplotlib.org/api/pyplot_api.html#matplotlib.pyplot.show):\n",
|
|
"\n",
|
|
"```python\n",
|
|
"plt.imshow(image2)\n",
|
|
"plt.show()\n",
|
|
"```\n",
|
|
"\n",
|
|
"Funkcja [`matplotlib.pyplot.show()`](https://matplotlib.org/api/pyplot_api.html#matplotlib.pyplot.show) posiada opcjonalny argument `block` (domyślna wartość `True`), która kontroluje czy funkcja jest blokująca:\n",
|
|
"\n",
|
|
"```python\n",
|
|
"plt.imshow(image2)\n",
|
|
"plt.show(False)\n",
|
|
"```\n",
|
|
"\n",
|
|
"lub krócej:\n",
|
|
"\n",
|
|
"```python\n",
|
|
"plt.imshow(image2)\n",
|
|
"plt.show(0)\n",
|
|
"```\n",
|
|
"\n",
|
|
"**Uwaga**: jeżeli w skrypcie ustawimy argument `block` na `False`, to po uruchomieniu skryptu najprawdopodobniej nie zobaczymy wynikowego obrazu/diagramu, ponieważ skrypt zacznie wykonywać dalsze polecenia i np. zakończy działanie programu. Ustawienie argumentu `block` na `False` najlepiej sprawdza się podczas pracy/eksperymentowania w konsoli interpretera Pythona.\n",
|
|
"\n",
|
|
"## Moduły\n",
|
|
"\n",
|
|
"[Lista modułów OpenCV](https://docs.opencv.org/4.5.3/) jest dość długa i zawiera pakiety algorytmów m.in. dla przetwarzania obrazów, uczenia maszynowego, głębokich sieci neuronowych, fotografii obliczeniowej, dedykowane rozwiązania dla CUDA, przepływ optyczny, operacje dla obrazów w logice rozmytej, przetwarzania RGB-D i wiele, wiele innych. Z poziomu Pythona dostęp do poszczególnych algorytmów odbywa się bezpośrednio przez moduł `cv2` (u nas alias `cv`). Część klas i funkcji ma przedrostek oznaczający moduł, np. `cv.bgsegm_BackgroundSubtractorCNT(...)` oznacza klasę [BackgroundSubtractorCNT](https://docs.opencv.org/4.5.3/de/dca/classcv_1_1bgsegm_1_1BackgroundSubtractorCNT.html) z modułu [bgsegm](https://docs.opencv.org/4.5.3/df/d5d/namespacecv_1_1bgsegm.html) zawierającym udoskonalone metody segmentacji pomiędzy tłem a pierwszym planem.\n",
|
|
"\n",
|
|
"## Ćwiczenie 1\n",
|
|
"\n",
|
|
"Wczytaj plik [`img/soyjaks.jpg`](https://knowyourmeme.com/memes/two-soyjaks-pointing) i spróbuj odtworzyć poniższy obrazek."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 6,
|
|
"id": "5F_V30CUy-i7",
|
|
"metadata": {
|
|
"colab": {
|
|
"base_uri": "https://localhost:8080/"
|
|
},
|
|
"executionInfo": {
|
|
"elapsed": 17024,
|
|
"status": "ok",
|
|
"timestamp": 1666007708904,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "5F_V30CUy-i7",
|
|
"outputId": "0faa68db-5115-43c9-f9b2-43aa4a771c82"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Mounted at /content/drive\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"from google.colab import drive\n",
|
|
"drive.mount('/content/drive')\n",
|
|
"%cd /content/drive/My Drive/aitech-wko-pub"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 7,
|
|
"id": "PCVjMgv_zB7m",
|
|
"metadata": {
|
|
"colab": {
|
|
"base_uri": "https://localhost:8080/"
|
|
},
|
|
"executionInfo": {
|
|
"elapsed": 242,
|
|
"status": "ok",
|
|
"timestamp": 1666007758111,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "PCVjMgv_zB7m",
|
|
"outputId": "b2618783-2585-48ab-d1df-d5ec484ba547"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"/content/drive/My Drive/aitech-wko-pub\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"%cd /content/drive/My Drive/aitech-wko-pub"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 48,
|
|
"id": "ff692a8e",
|
|
"metadata": {
|
|
"colab": {
|
|
"base_uri": "https://localhost:8080/",
|
|
"height": 265
|
|
},
|
|
"executionInfo": {
|
|
"elapsed": 700,
|
|
"status": "ok",
|
|
"timestamp": 1666008786851,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "ff692a8e",
|
|
"outputId": "743a4a6e-9786-4001-c15e-bab7412b03f7"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"<matplotlib.image.AxesImage at 0x7fd69e926b90>"
|
|
]
|
|
},
|
|
"execution_count": 48,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
},
|
|
{
|
|
"data": {
|
|
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAL0AAADnCAYAAABLwlehAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy8Z5Rlx1nv/auqnU4OndP05CzNjCSPgmXJtmTkbGPAEZa5OBBMBnN5uRfbXJt0uRi4RBsumGDA2GCDjRN4FK1gzSjNSBppRpN7OnefPnmHqrof9unuGdmWWa/Wet8PqNbqNXN2qKpd9a/n+T+hSlhreaG8UP4zFfn/dwdeKC+U/6/LC6B/ofynKy+A/oXyn668APoXyn+68gLoXyj/6YrzXDff/X03W60NBkOhUmZkbJyBgUGGijEVp84vf/RfeMUNe7n2QImbX7yJi2emWWoJPvGF03SSHMrxgIRdWwcIHIOSoA2EcUKUJERhQqcd0mpFWAEmtvRV8jSW6qws1OnrL2KtwRhDnGiWlhq4ysXzXUJriKXCDQKk79Bp1+nvKxFFMVGs8QOPTiR46ukp9u3dQiWTIHWTgWoOR1muu3YzO3YOIAQIIZBS8fSj03zmLw+z/eo9GJPQaHWQjqXdDEnihKwHr3jVNgp9CmMtrYblK/92kkbbstxI2DE6zKaCS7FaQkqBADrLNVqBy31Pn6EVW+LE4vsBy7Uu1aLLno0BrmOZ3FLlums3gzVgLYtzbZSAcl8G4oR2w/DwQ6exCFxHUswXCBPL/EqbC7NdVlY6fO+rNlApSAhb6G6HQtlHOF3QIRgLQoCSgACrwNr0/wDG9GZdsLSYcG4a7niwzVNPtLCdEEdYSoMBr3/nDWRzq08K6I2ftZYwbqNNjLUGYdPraZFYC63IMLOo+OgffoPFxXQcyvkCvh8QhW2Gx7NcvHCK5YVFgkKOlUadUjHPq64f5ud++gb+/O+e5oqD34sUkm4YMrewwJOPHqHTaSGlwlqDtZbEGP7PJz4v+DblOUFvjEFrjVSSqBsTdhMSbegmCi/nMTJUYKHWYG5OgJC4jsJ3BUPVEtPdMsXBUURUZ2l5loXZOYRyqDe7JMYgAGMt1gqEkNxwzQ4mJiosLzWZVoLmSgcpXfzAx/MDMtkCm3bmGOjvZ3RsiGJfiXy5gBdkEFJy8sTjHP7GnSwuNllptLEGfN/H9Txa7YjBYo4kaWGtwPMcwihEKJBCIKVESUm5r0S1XGDLllE63TYnn5mi2WhjrSbIeiip0driKwcD2Cx4gQdhgnQMiYQ4CvGzAWG9jTEJruviOi5GCIw1GK3BWCwQxwYrFMiEONZYIVMIWktff5AC0egUWMoiBcSJAVdhhAEJ2hqsNXS6EXFkMdrgScjkPKyOQSRgdQpQm9adQl0Dsgf89WItJInBdSTZaoHx666kO7vMqSOHCbRFG02UWIw16ULqvW/QxDoCAVJJhBRI0sVghUEbSZRoFuZb+CrDyGARR3hIJGE7IYoFRgscx0EnMcYY+gcGyWYCiqUSGokQLvTGUWuN1YYoiojjGCk1cZIQxunv5yrPCXqpBK5Q6MSQRDFRGKITTaxdtHDYtnGYx5+cZttkjjAyKEcSeDA+NsCFM3DyxNMM9+XJ6y7XH9yJ5zvcf+Qk03N1pBTkggy+5xMEOQaHRhkeHmZiImDrdot3W5ZSuUy+WMQPsri+g1SWKOyyUq9TX1nhzLmzXJyepdlo8epX3sbU6RN0G6eIPJduovFcSSGfxwqJG2SI2gmJSbA4dDoRWLDCggBtNX6gwIb4vqQTJfiBorFscFwHISxSOiTa4jgO1lrwJZ7vYqxFSksMZIoFdJKgdUK2XEQkMd0oRrkuKhFYJXqyVRJGFoRCSIMxFislwgqwek0AIwApcDyPwY27ME4JIWLaC6eJkg5xEpEkmk6YECUaoxVWWLRJsEkXVxpSgWtBpyAVQoBU6TXE5cC3AovAC1xmlxb46oPH2bNrD8WxYYzToR120coihERikDYVHKvfZDEYa9PfAow16ULSCXFiqC0nDFRHqK0Yum2N1RZjHazxWZwN8fIlJnbmcTwP33XJBz4jIyUSnSClg0JgbPodNklIkmRNoygpyfoB+MHzAL0EK0Abi7UaHWuM1iTGJYodtm8c5dDXj9ONoLbcIR8EtLtt+voCMstZql6BYlERtDrkcy5SSW688Xq0KFAtV6hUquRzeTKZLJmsi1AWrTVhFNJstliurXDi3Gmmp+eZnllkfm6BpcU6C8tN2p0ucaJJEoPAsGP7Xnw/T6JTDZXEMfmiS6lcxCJw/AAlFcYkIATd0AAKKdMJF0IQ5BVBTuI4KSaiKEIFHoHvYXQEQDfUSNVbBK6D57nEOiIxgnYnAqdMt9FCCYG1GmsMnupplxisEWhjEMYSJQZtBAiB1qlqRtIDoUzpiAWMxgpLrEO6UYNM4GKRWEBKBykNsTYYLAhLiokekI0A5fTqNGB69aa8hFTdkf4L2AQcqRFSEsaaVjvi8COPs3vXTjLlIokF1wBojBD4QQ5HOoSdFYxNF5cw6Xjq1QUgQGuNEA5T50JqC4Z2VxIlFiVUb7EohLRU+wpoEa71P9QJYdwhSRJcN+VVxqSaRhuNMRrXddeAb0xKh58H6CXWGGRvkHSckEQx2mToaMnIcIEEaIUwO7NCeXMJi2FsKMuB0j7OTreRcR3dmcJaS6E0wo4rXoIW0G53aNbrnD9/lrmFBWbnFpmdX2R6dpl6vUOz1SEKI+JEY7TFkq5uY1kbXGtXNaxludYkyBapNzu4jourLRJoNRu02iE7twziOj6u4yKVT5SkEk1KscZ1vazCy/gkcUKQyVJfbuIFufT7lUunG9JoxEjpAAYpJI4raYeabpxQb4bIjE9poEBSb2PCmO5SDaevhOs6CJGkGLYGgaUbRSRGI6XEJCmbWaXcCEmPjwACpSS5QgEZ+iRRE21BuRl8D1wPLAJNACqHlgJrJVI6GD8A6WOFh9B1ZPtxsB0wFqFk2o50wfHBkUgLJTeDWRL4/gqel6Eba545fx6t81gGUq2ExVEuTlBIeXS3hbEJ1oChR22EwPQ0pOem0ry2PI8j8zjSgpRYq0iMxApI4hBjNNJd9684SpHNKsKuRkq3txgsWIHWOh0/k1JUa+3a3/9r0IteRakG1ERxRCeMyCWWMBH0F7IMVPIsNLpMzdTZs60fKS2+6eLJLlt2bGTnpjHu/KczdOKQsVyR2flFfvcP/oLZhTphNyaKIqxJJUKPcoLtzbywpNBMV3A2myUMI9CaXbv30lipsWP7dg7dfojlWp3NG6rU6136+wtYUq6bDTI0mxHGOjheHulIHNdFG4ExAsdxsUKASieqWM7RaYU4eQ9tLK4jcJSgE1rCWNNsxek7VoNUlIp5lGoCmjCOSVY6zJ+fxbZDkk4HoQSlUhHP87C2i7UWYQVSCRKdkA6vTLUpMpW4wgOpsEYACqsk1jpYIxC4eNkq2fJWHOUzZAVbrcuNt0iqlQLKdQGJFRKNIkktdazVqOgCbuccChA9CYtQqSbwnFS1C4Hje2S0pFQoUi6FNLshpVKJSqWIUD5O2gRKSOJuh0THaAuxUbTiATo6j5cpkstWyRQqCMdDSJdoKaTRXsAaB1cJdGJJtEuQHUXgYuMZhF3G87r4vofn+/ieQ18pod1JkK6/BmqtTcrdjQbhpnjtGdTfqXwH0AvQYI3FWEsYdul2QmJt6SQKIVw2TQywsNhibjFLpHt8y7ToLJ3ivjsfwdx8A0a4dNoNut0Wfsln6mKNeqOLkKnqszadhLGhYRYXFzlw4CqGhgbJZDxmZ2fp7+/n7z71D7z1rW9mZnqK2+/8Oj//0z/KZz7zKd7+tjfz6CMPsbRUY8+uTdRbMZU+ief5CCHJ5zM0WhFSOfiZLEqltEZrCMOEbFYhXS/lccJSKAbUay0Gy1nGh8sEnstKKyLIeLQjTaPZASFQUvVAn8PzPCJtSeIY7Tp4hRzFrRvIDVVwi1ncTJbKnY/SafvIgiQIckgng7WCHTvH6CvnyGUCJKMY46UOHBQGB4SDkWCFoH9CsUpL7JpBuu6BsasakNUxTX+nT0msCEBmEWiEkD2NIrBCIYRKbYoeTRAShJJIJVFKUigUWG5ojhyd54arB5HGoNGQaKzQCGEJk4CpeoXpxRghm5hkicD1UEZz5tgRVpYDlheqJEmGRCcYq3CcYYzegu0UUFykkDlJsa9GELhp+zKhkNO06jVcv5zSF61JdEK726XnQFoD+/MGPYIeT+zxpSSi2+0QJjFdnSHBZfPkME8cOkqzM0irE+I6ijCOGCs5VAsVjh09QWAN0sa0mnUmtmWZnBjl0cdPcP01B7np5hfzsY/9KdZY/uvP/wx//LGP8yPveTvGdJi6OM2Z00/yytvexNcOfY1KucCBK2/hgQeP4LqKuZlZtE4YGR5kudYgmy8hbIY4dsjlPISj8AOIo4jFxWUqXpvYhjiyiKd8wk6CLTlYoxFKIaSg3F/i3KkWg2KI0lAfK3N1lhdXUJks1kC7Y0gSD8eVWDyKFYdNkz5uUMDzPLYe2Ea1UsBxfITjIpRC4fDdb7wKi8J1HaRSSClTY1BIrNY9d5tJDT9s6h1aBfPaHKxPqFiVFpc8Y0U66auuxPRuT/rZHlkSqfvQ9nivtAqEwSYKIXSP5wusFTSbbZaWlkmM5cK580gpiJseV+yqkvUtVqy3iTWcPt9lMWqTKw0RxjFC+QyNDhHV5qn29XP+aU2nsZFYK6xIUG6WON6A7uxAhRWE00+3NY9qXkAnEqUUrjSceLrD+fNn2bR/V48epraftQkC0YOnWB+X5wP6IFPAUMcKgY3SCel0u8RxQpxAKxaMj/XTaLWotS3zS20Gyi6ONORUzPSFC6w0Y8YmKsS6w9Y4pNupsXPXNh499hRx1OXG66+htrzIJz7xSbphh77+KtOz89x5190c+tq9xEnMrbee5aYbr8VTgpGRAbZv2YQxCbVandnZeSYmRpmdX6SQzRO4ZZLIx6s4+DkfP+ewspKglEeQVajIoIRESpdOaBEyDyIA6YGTYWCjy8LiRXLlTfTZIQZHPHa/KIOfzZHNZ/F9j6Cv2vO6OOzcJ9mxL6UJQmik6Fn/qxNgDFhwvVQmp7bmJSA2OoW3SF232vakuLFrPvDUzWgvU982baBn61w+b6YHgUsl4GrphjFPPjlFPldAWIOQglDDUiOmUlVcceUgYIm7Cowhn82AcnA8D8/10FLy5DNtrtyRQ0oNQKIls4vw2LEamZJiJJuhr5BjsODT517ELWcp5a/h3CmHU8fzOJ4lW1KoICDqDGE7JawOkJkcGS9LMZfBkvRodcJj5zqcORez5dpsGs8hXe860Wit1wD/Hy3PLemlRAiFUqBVOhFR0qXbbmMKebqRZKCcJ+t7LLUSzs+2GagUEUKS9RL2XLGVWtvDFQnhzDxxHFNfusjevdv4x88qLk7PsbLS4MD+K3j8mqs4f/4CGzdOcu7ceV55223s3/ci7v36PXzpS4f4/nd8H/WVGo888hi33vJSjNa0w5gzZy6wcdNmzp2/SNbPkS9WMSbh4HU3sTB/gSiBrFuiWi0xNlgk41hK5TLFYpHqQAGCEla6PR+5x4bdgrGdoFRKJay1CJkaVqtDu0otVuXs+pBLrH3WBAjZCzitO016law/0rtmSYEve56WSw0zu+pqZBXwq/X0pPzq/Z7Hx16qFXr/mkRy7lSdu+6aYrBcRDoGV0kcR1LvWsKKYiDs0O0Y7n0q5OSMQ7GvTCaXJ58vomRqa11YSL1bo4MeBjg3HXL6fJNaI8LYFVZmzhF0QC/VqIznaa9Ijtxd4+Tx/eikgBWG5kobvdzAJJqo1cUYQSaZZnhlkdKwJUwiwFIIJEGQxfN8HDcNqBlj0NYSR9Ga+/iy8Xk+hqxSLkoqrEl9vVaATTQ6jtHa0I4l1YxDf7XAQq3F7HwLubMKNsKlS7RUI4orxEJQ7huiqwXtZoPJjdvoq5ZZWq5RW2lw912388Y3vJannnyajZOTHDn8ELe87MXMzpwFpTn2xHF0YtiyeTOf+Mu/4s1veSue7xLFMU88cZQdO7cSRSHSEfzwj76VfD7Hhk1jhN1W6kOWAsdJKcUq+oQVpDJDriEjBWxquK6LT5u6/ejR/jXSwHqg5xL1ug60VfDRa2+VjayR0MvfsRaEXIsQCyF6Ktxe8ohN6SaXdI9nqfRVHzmXPNajOtq6PH22wctv2MR41U+FmZWEMdRaEbGGo1MRJy50ueuxGpFfJSGkrz8NElksnpehWu1jfnGR+aWYVhjRbIVobdIo9uI8C089weNRg6JjKL7+OqyUdBeaNOtNYltDxBphEywJRncxzgLKdpHuLJ7X5uCejSRxC09Zsq6mWV9gYU7iOc6aO9YYQ7vTRUrZm9d1sK/+/nbluQ1Z4SKVB6sBAGsxiaHTCdEaosTFEjExWuGxUwss1Mto4+I4HsZaNm8a4UyjShhp0A7NVp0kTnAdyfZtm7j33oe4eHGaMEq4774HeOV33Yq18KW5OeYXVvjTP/8H2p2Ebphw6Pa7+amfei/LizWOPvoY115/DVonnD43ww033oxUCoNh585JHCd1KWYymUunnvVgTO8PdRk6LrP+e98rLgMmiF7U1K6B9nJDUvRoS+ppl5e8eCkMZcqHL5FOq8bppe2tuuFWJ/Rbya/L6li7tt6OQGBN+rujPT59zwKBE7JhOMfySoelekQ3hlo7xhiRenWEotU2jIy5tFtNfNdFWIMEihmPLZvGaTRWOH9hDmMF3W6bJIlIopBOu47vKOqLHbxshuMn5xmbHKCvXGKoTxMVm8SJpd5pEiUNsCGOiAmcNpVyk8nhEnsmNmO1k3rITItFJcl5EiVSAZz6qg1JHCHEN7sqn5ef3gqLFIokSUgSTZJolBKE3ZAoitCJiw41m4YrfP3hszTblmZbE/gKG2syfoJtSoJyiUBlCHSO8S1XE2RK7Nt3Jfd8/SGeOX2OjZu38Kcf/3Ou2LubAwf2EycJ2WyBj/yPD6KxfPjDv8F99z/Ia1/5cqJuyKF/v50g8IgiQ72+wslnjhMnMdauA2XVcSF6nGIVkGv2H9/G6FlVkcZ80/1nq8/L1Ki4pD4rSNn25VJaXMZv1q+v8vvVRXdpu5cCfzW6ebn6lin/XyVa635fVvUSvUsiShhrxpyqNTk0E5Nou6ZwfAX7JwqUM4qWURw+EVLKOoyU+3njtdsIXIUUgihJOHfhGFFriW7SwQpLQhvpGXwXwkQTL9fpz+apZDymphvMNSJKY4NM7j4GKkMcGU6duEC9USOO24z2ZRnty7Nl4xCn557kTz83TdiNWKk1UAoazTaFUt+aC90Yg0kSjI7XtKLpzde3EgLPLs8J+jiOem6jNLgge7knstFk1z0PcsX5ixSjNlujkPFai4WNyywsNBkZzWKxVPOKvmXJ1f/6ryzlXGovPUi1fwKAnTu34wc+J55+iu1bJ9i20qTygY8w9+PvodFq8alPfwrHFdTqDXLLS7zmmXMEb30XP9Vs0jWWk/c8wFClylQG/vlfvkYu610Wkn52WcPjswD/7EF6ruH6VlzxcmoBVmvEF/8dnjwB7/l+bLm09tylWuSyBbTuX/ymdtZciD1DV6xKsZ7Bq+4/gnP714l+4M2YseFLe3ZZXTYKqf6vP+Ajdz1AHfg/V23n37N5+lpdvntukemNFYa3FZFoTszHxHFEfWWF6lgf07NTxFFMqDUz9TonFxcJEURSI7EIJ0IKDRZK5Qzd+Qa+TY34qBsSAt25OVAhOCCsINYzeK6hmHGoFBSuY6mtdJhdbBB6isRo5pZXGBsdZn5piXJfBmUMo3fdj1hc4htX7iJJYmRvXJ9NcZ6rPCfonUweB0MUhkRRjDBQabd410PH2DO3hOo1UAG+G+h+8QjzBYl450G0tGRFzKbPf5arvnwHWknu8BX6llegpMPo0BDjo4Mce/w49RPP8BsnTjLZ6fDE+z/AzM6dnDp7ASxc0Wrx4VOn2dTpXgbIq1dWuHlmnr980bWYW1/O3iu2UamUv+1KXzf05LNl7dq9/4iU+Lb1ro7106cQ7/tFWKqlE/D+911OSy7TFpe39+2m69lc1eg0Ms1ijeBn/jvymTOIqWnav/uRNMD0rPcsIGfm8D/3RWQUEwCvbTe4v1TgB09f4NVzi1ycXeC99Zh5BJ3IkGhB5+ICoQOnajXiOEZjkb6D9F2MBaUEJknQnRZSgCsl3XqIMGkym9YWEyXgKOK2YXikQLGSw1qH0ydmWWqGSAGhtggvy4XlOvVuTDS7wMbJKpWNVZYW60RhRCYIqJw4xb6P/xUyjlkIQw7nQDhOT6t+8xh/u/LcnN4YjBBI5SKkJh+HvOfIcfbM1QDQSpL4HsoaVCciiDQTn3qQWimA79mP9BwqMkJai0g0W448Rtjp4OcLuEGGXbt3cfLMFFvm5hnvdBBAOYxwtMYqxbZWi185eYqxMEwnL/Ah8CGKod1hOAp5//FjhD//buTLr+3ZiOI5pL1ASLXG163Ra2C3Pd596eD9RxaDeLaEXliEegNhLfbMudQuEHJdmq/aDr00ivTSav2pjz4lKuqyPkCPsIg0mgtAs4lcWExz0s5dQGiNvTSRrPeWAGylgt69E3HP/SRScGdb02p2GQ1DBJDvRHRqIcuuh9YaRzmEcYKDJesojBI4SpIJXBxHkclmGKhkcUmwSQtXKZSUnH9qljPTDRISXCmJIoPru+T8DGPVAqVKhpPnFonjhEgnyEQSC8lsvUmt2SZOEujAxQtzTIxVkTLB8yRBJoO/VEP1+pudm8ds6AMXel4ItNZp3tXz4fT0ghqo1PC64cw0u3uArxXz3Pv6V9DdsYEhZ4XSFx5g59efxIk0hU8dpnP9ZsygJKMSet1i5ORpZk+fhiv3AZYrrtzLV774b3zX4tKaSWmBIMhS7uvnPd/4xhrgzbVXE//yz2HHR5BLyzi/+fvIr96BXFwi+J0/Jr7+ACKbWes2EsSzPBzWWnQv7TSNuK+6GNe57RpIBD2Pi12rcw3fvR/fOux9CTWxvTyhXkcE0I5WmO9Mo7VhtDRJO25xfP5ROqZFZLqEcYgUirJfZUNpMyOFDShcVmtYde8L0fsILl10q6tWrD2fpm6DLRRZ+p1f5d9+8heYtQ0GXr2N90oYu3geVppIAf15lzjrs7TSoVzwqDU6NJY76DD1x6+mRSc6QWLwHIE2CWE3QuteqnE7pKoUYIi1JTSW2vQiK2cvcmF+gW07Bnnk8ak0f8dapHJQnoeVAi0giQwKSxQqZmdq9PXleMWNW1C5ABuvj22SxGmejhAopS4DvLbPA/RCKIRIU2kzseaGM9NIIJKST193gOVd2+irlPGyeR656QrOPnqK1zRD1HwT99CTJN9zAE8ma/UFjRaVO+6ju28/INm5fQdXF0rsazTXpq6vWuVXP/BLVOIuY3d8Lc27GRsh+v3/id22MTVmNk0S/+9fx/ueH0QcO444fgK5VMNms+vgWN00gcCK1JNirU4zR5stuPNe2LMTsVKHO+5FzMxBuYh58UHstVdhXYWhlycjFCQJ5huHkXffDysN7GAfvPQGuHJv6rS2FqamST71j3g6BQlHHkX+rz/EvON7YGwEaw0XGuf40snP4cuAN+59O13d5cHp+4lFhKWXKYlAGDj56CFuPu6yeTkLnoc5sA990w3YXJaeV399suIY59DdOEefQLTa6PFR4pe+GLtpE7YXKzDlCgvFIscWmvSfXmbPaG5t0Xqu5M0vnaBWLvCZrx3ntt1jLNz5OHedmeOYo4jDEK01OWN5SRRxII7IARd8j7szHtO+j7GWHJr+nOKKRodF3yM2huuXWwyjGShnOPnMAoebIY4SZHM5xiv9vG52keGelO90QoQAzxVICTMFjyf3SMb2bsVGl4I+xdUq5VNK4TgOruuin4+kj5K4N56aiaU6QystLHC6WuCR/gJDYUxiLKFRVIYq/PNAiVtac2QsZB84g3nVrh6/Wy+5r95O/J4fxObyDFTL/NKubZQOHVq773kOExMDeJ/+LLLRTP3Lr3slZtvmdDfSqjQeGUT/xgdQv/o72GuvgoEBRJqXi8UgpEIoiZVgre7t6tEkOiL72x/D/d0/g8EBaLdheWXN+ahyWRo//nbu/t6ttG3MVaPXM+70IT74m7if/BdotteeteUC3fe8GfELP4knM4if/mW8rxxa0zDiiafhiY8iH3yE5K//EOH7uI6PRRCaiCgJ8Ry/12/IOiVcmUEnERNfe4Sb/+whSrOtNeWhXBf1kusIf+tD2I0bLrO61WNPkHvvT6fUD0AIgpEhWh/9deKbXoxFkv/CV/iJ2+9h3lX8Sg4+f3qRa2pdhoAoNvztV07wjHUorXS45fAhdndCDmeyvHtiA62sy0Qc84GZaW7qdFJWQSpA5lyXv9+xnfs2byAIPG57+hm+9+wZGkqkWZs6dXfahSYvkYJqX56/2DpOqVrmzc+c551nL1zqPL6s6Bn4p4zPqas9bDNcu256QcNV58Xq4v1OG0jgP5JlCfh+lo2NCE+nK+joUIW21XTabWKdEFpFoZBlZrDKuQtL7IgT/KkVConCc9c/RwDq+NOoR48S3XAtqtmi/+6vf0uPifvosVR6ui7JrS/v8WDAWqQQdOIG3YMbST75YRJl0dEZchQpBwO4joeQgiemj3Bq5WlacZNO3EVbDUnCqw/9K2Nxgp2aXmvPSpGq51ab4H//JbWJ22js24lvFebnP0Dwt19YzWlee1bUGgQf/QsW6osEH/olcpsnEa6HjaK1hWE8RbJ5HCEVFoHvZFHSIbIhnSSk6hWRQmEM7Bu8lr39B5Gf+Wfyv/unOM3u+sClM4o8dDf+e3+W9if+8DL7Q8TxWiCM3jiJizNkPvI/Sf7xb7D5PM5DjyKjmP4opnWqyVPZHN14PcbQSDIMIfit5Ytc0e0BTCncTJZJnfDbUxe4qtNZH7Ne14bimB95+gSFLRM86Ss2z83jWks1We/hamqEayxvXGhyeovL7UpxTkpaSpLX6W66NJC3/p4UgkIhR+D5YLtr1xyZa7IAACAASURBVLVOekJw3bsFrG/weY7yHRPOrDFYqQhWgzLAiu9hjSbsdtCJJtEOjq+oDpZoyF4nEk3SjUievfK6XfwvfJnw2hfhPPwYzlMnvrldaxHd1FtjlcJWKz2DUKCkgzKG0+cPc2zxG2hpCQMHLUFJj/HsBNdOvIyCKjHbnOXx+eNYYVGOQiJQ2pLYde2TeIqjr9lB5tZb2f7AReRffwa3E7P5vvMs3nQbzl334f/Dl9ItfsU8rXe/FbNvL/59R/D/8jPITofyX3+Roy/dzbYP/BRq0xjZX/w10Iaz14zyjXfsZ/yWG7nKTUNVjnLBSoy2hHGIEgolXIxNN+h4tSbZj/4ZstkFKTn34m0cfukQhVrEDZ89Qe7CIvLho6hvHEEfuHJ9yAA7Okz0I+/CjI3i/+4foY4+jjpxCvn4cfTBqy+L1vpeQJDJrUcvhWCT0fzE9EX2dlNwnQoy/M7kZvB93jI1x4FWC4AVx+FLfVVWXI8bl5fZ02rhRxE7/+0O/qlaoF1rrLWTCLi7r8CRsTIbF+q8aWoFx1quujjL/VcW+GpfganMJK8bLzEeNdl6cA+Fv7sXcX4+XcSvehHxa6/DCQMMtbXv1UmyBu5nu5+fXz69EEiVpp1Ktf6o6BlIcRQTx5o4cTGBy4bxvkvfxhiJ6L13aTf8r92F8745/M9/EcJw7d5z+8glJrHos6fJfODX2PXYw2zWHYwjOfbK7Tz8lqtpqTaPrDxGrdHi1i2vI+dVUdLBdQK2lHaTc3J4RlL2HgAuglKc+ok38rVXlSlmPEqbXs7IP3we0Q0ZPtMiKIyjPvt7iCgBJYn+208h3/NOhJCY17yKJFfA+e0/wm1HmC98lTPX38CWvTtSA1MbGoN5Tl3Rz8Wlhxnu38ZobhJXuLjCpQO04xZSOjgyABRhEqGOPYI8dTad2Juup/77P8PJmS+hbULl5ts48Ce3Qxih919x2aDa/j66H/99koNXgxDIEydRRx+HMEQ+cxpz8OrLxlMpp5dflJbAGD54/hyjUbpD7HQQ8Iubt3AymyGvNS9bXmZ1W4tvDC9bTgHo9/izADaFMYFZ75YBPpFx+JuhIsWhfs71D/HKuUfIxwmTjTZVkRAXfR5sNhgdyLM032b07qMUphbSOnZPwG/+II17nsB1vTVNm+LBIKXoLdpVcZyW75SG8Jx3V6NdUkoSz1+rttyNMDZN4o+jGG0lkVZsHqzQt7rqHIXI5C5rop7Ppo1evEjwmc/h3X53+myxiM3nLm0Ym8mk7ekEsbC4lgLgfeUQ8gtfJjg3S2mqTvlsjRv++jHe0DnIjcO34RJwduU0080pKvlBXOnjWo8XjV3H9eM3cc3IdeS9YtpONsPgW/8Lo8VNLEdLHF45iul1t+QWyOPjPXai5/IrY1/5CoRMI5M4DvHrb4NsFgH0P3mRp2ceI00KTktffoiMLNJKWjxw/m46uoUQDkq6JFg6SReFwpUe1gq6cRf5yFHoacfkNa+kf3gbGSdHLBLmtg3Q+Zs/ofPJP8NObuzlD6XFbNuCvmrfOr0pFNaH81sE7VZzfFZFjWstoz1aFgnBb42P86TvkuiIcthhLEylvwAyxjAcRQxHEZUkWYNcRwr0Je0kAu4d6Ke8aTPZcj8dx0k3tQAyiWmt1Ik6Lcolh0bUYuNsk/4HTqXgLmfRH3wTOm9oLMzjOu4a/tKg87pUN0avbRPUOiGK17n/tyrf+dwba7HGcL6/ROSkUnvvzCJ+lBBFIe1OmySxREaxrR0yHqdWdWeiStPTJEm01tHHx0cwvXNAsh//C+TMDADx9Qex1ep6kwjiq/anMxMnqK8eWsutiF9zG/oNr0FvnkRnvNRHHWsGRY49g3uZrGxEepKVcJm8lyPjZkhsTDdpI4XoeUdWLU0IZIbrNtxMVuVZiGZT3g/k3Dw20phVeqZUSrV6/TAWrOuvBYOE1sw1pmnFzbXvqAQVDoxdj8TlbPM0x+YfAungOTmEEHTiDliBr4L023SETdYNURH4ONLDdwPAECUh1vHAz/SM38un79spdSFEmnh3CSC1TWnealjn0gQJz1reMTdHOYrBWBzsGs+OgSnX4YLnpn9u+veU5/InpRItodZqEkJQHRggijXtZgtHiUtMDkvWhZ0bMvzQG3fxQ1urvPj0EsoArqL7IzeyMuawPDPFwuISylUsjQ7TqlYIfZ9nBqupC9oYjNEkSZIK4SRBm+emN88J+nSPLBhtmKmUmKuWUjVWa7JvZoE4Cum0uyRaY7qWzXccxdEprJ7ZOkQs1/NPLPBUf5V6XyWte7mWuhU9j+7rX4O9RNUiIL7pBmwpbc/5wpcRT59I91yOj9H52O/R+ps/JhlI67KBR7eoWAxn6NgWruORzxQp5aoUMxWEC6frT/Pk4sMcW3iARrSyjhIrGMqOc+XQdVi7PhwShfU9libKqcG2tIx84CGMFRgE1ljUnfdCK/VozW+s0iaio9eljG1H7KpeycbSDhKreWjmQea607iug8XSTbpYFK4KsECYdNE7toFKXaDq0F2YTi9gYwXezDL+h34D/7/9D5hf/Kb5WgWtsZfqm9WgnLzMxZmYdHvjpVHb867LlJsKthvqdd5/4Rxep8kCCXO9610p+fXxYX548wbeu2UDP7Zjkn8Z7mfRc4kyDmElhw28tTojo4mj9Iyj6YvzqT+f1MVYyAeMjffR17Vs/ftHcerpTqjw9Vdi3nY1NqNYbrZpdTRCKFaGBnjgB76PO153K8fHh3Bcby3L0nEcPM8j8DJkvOxzwfq5Ob3WuneIjiXyHL6xazNj80s4xvD9j57EM7DoZqkCBx4+yvCR1ChdzPocKge82IheMlQ6I8uOy6mdG7lqbnFtUJKtm0muueqbCL2enCR6xcvwP/1ZxMwswft+ju5/fz92fAyxsIj/W7+HOjebTtauPm7nXqIpSTtq0Rf0MV6axEQGG0pWGnUe6H4dx3VQ2jAc1Sj3OmWtRGjFFQNX084cvkxcOm7AyRdvZNM9JxBxgveh38Rqg9m7C/XAYbzf/oM0Eugpnrl+EgvE/VVsIY9YquE+cpzMoyc4OLqP2dYU9XiZB8/fDSrVOGHcxVrwVQaJJNIxybXXYEeGERemUP/6VeRQhvzVltJyjQN/fSfe0SmQkuTg1Zir9n3TnF2ixy4rz7aX4jgm7HZ7mzCgIwS/MDSIJwQfvThNVWtes1JnRgn+eLCPr+cybAoj8sbwrotz/F5/mXYux6trTd46s4Bv03jO01s2InoBTCzExhDFMd12m0Krs7bIGp2Yrxw5j3zwLL9da+G3U2FxxlXcpSSlvz3CwnyNMw3NuTjDLisZf/go13zy08huyPLOSe7avx3E+r6H1YCX0c/DkNW6ly3YS+g5vHWSbSfPcMXUPJVuyHsOP0H8yFNIBG6PBnSF4GOVIucTyf6uYc1fYKHR6XJk9zh7H3gMP0yfD1/1Xdhi4ZuNWNeh87734jz8KOrkKeTDj5F927vAcyFOoBepbQ7kuP+HrmHZCYmaEcYkROEcD576OvsHDlL0y7iOj+e5eDJI06StWkOCRmAATwaMFjesw8NCXpWZuXEPZ79ylI0PTSPOXyT4sfeD70O3m4b9hWDqpr1MXTGSHsexYSPx/t14h+7FOXkG5/XvYMO2zRz84Fu4o9DgXP0MjpNGIEMdYa3FcwKwEJmIaLCP+N3vxPu134Ioov9PPs3bAwepDSpOpaQZH0Xv3smznRSrXPc/knSlE00s47U03FgIpqViJpvlI8PDfHhmhrzW/JelFc739/PliVFu6UaMtDvsDyM+dnEeLRYIegd3JULwWF8J17HpSQek7LTiGUZzAreQJ5tRqDMzoEFKgeP6vGVphVvb4ZpdMBZpvufTR9b6Oe27/NVLXoS0sOWe+8ktpQvq5mcu8OCeLbT91MZaDcKub6389uW5DdleHsuqld8JfP7qxv08MjGEFgJlLUGc4PUAH+Uy3HXrAT6V81lpxKw0QtrDRawURBkPuaHCMQcuTA6nPuz+PsJXfxfG90kmJtJA1MQ41g+wVpBs30bzDz+K3rs75bhRhGi2ED3AdyaH+Oov38rClVvZkb2WA6WXs6twEBPBY/NHmEvmGCgNIAXkZJmXjb+BV295B/1X3JC2PzQI5WJ67AsSb2CETjWXpttv20zgFRmavJIv/MJNPPOisfSgpiRBtFop4F2X9htv5ch/fT1hoMg7JXK5fk6/67U0+1IVK9odnKNPsvu0ZmNpB9pauiZeC1Bpq/GdDAhBbEKMTGj/8NuY+4m3EWXTEJDXTXBiA0Jgtm+l/Ue/Q7J9K6ZYwAwNpuO2eRNGqTUprzdMYH0PfB+9YRwrBHrLJqyUND2PaGCQ6tAYS/3p+7VshoHtW9ixezvH9+3lzzdN0hUCF9jtK7JXbORj1+3hQjF1OPjWku0BPnIV/7ZzlNMv2sBVwx7eaBELNB1FvhywrS/DgYkye7cP0s6nVK5eztJf9NlhzVpgSgAekLfrf1u6MdcZAVKwMjSYnlwBzBVzJI6T5jWRHkwmZG/nmXxu0KsPfehD3/bml7/0zx9aNX6kTI2htjA8PNzHVNYnFIJaxmd5oI+pfXs4+uZXsHjVRg4ffQbXC5gYzFLaN4EpZznzkl0cHelH5jOweQPSOLi/8gHiA/vAdYmv2o/N52n/5I9hBgdXTSHs0CDxq14BQ4OQzWDHRjHbt6Hf/CZmP/Bj3F2YJ44tN2y4hS3lLUxWt4BImAun8aTPUHGUZxZP4EiPq0aupRJUEQf2Qz5H9L73YLZuRkiJsYaWn3BPdY7OcAXv538Wv9BHwS9yIjrPY9dWqY9WyRb68Sc2Yg9eTf0nf4D7v28vR6NnAMnOypVsLG/jfDnmzg0tTBBQGd6E/u43kLzz+ylXRjlde4awF2RxrcfuwStpRiucXT5JTMh08xzHlh7hwcmIZ/ZWSTwHrzqEu+dK4re8ic4H/x/0zm2pEMgE6P1XYoaHCH/8vdh8oSf9BXZiAjMyTPxdLyd+1W2gFHrndh6/OMPfl6vMb9tOqVrkqKeYbzf45HAfh2sL6HadudmLHNaa2f4KU5Uid28dRQYOSSlgbvcwSUbQEdAsZjgxWOJz24Y4tWsMR1h8R3Eul2G2G3FovI+zhQz1dsiF+RWm611O5jO0Mz5f3DRISzmcLebwYsOCEJxxFGek5KyjOO86nHMUh4oZ7juwm5Et25ibnGClVOTpgTL/uGmIKJ9d2yOxumjS/CTL697wtl/5drh+TnrTbDcRQhL4AdL0MgKFoOMo7ts8zt1jA0jp0D88wYZNk1TyDkWnzkA1T73ZYW6xTbx9gKnX7SPGRT06BwhaWZ8HDu7ntmtflKYLCDBbNtH8xZ9NJ3O18zbNOTSDg4Tvey/Rj747DfFLkWZ/6hX84wFd08XKKNVI0pLN5FCuw2J7mR0qgyN9QtOlrVfwlUM8mKf1/vciEPgy3YBgrSXnFVjYP8nU3hFel1cEWlP0+rh24uXcrr/AkVd6PPGqgKHMMFkvw3JcY7Z2gsRq+v0xdg/tQ1hBxi8ws2OAlZ2TDO54G+VgCCkEA6LANSM3cMe5L6OFJiEiNhGe8hDCEtuEC/Wz6aYIa4n2TuBefwv9ozej/L61c2lW0xwskFy1j6TH7S8N1ljPJXzHW9avW4vJZfnKvj185uTnsI8vEUcdXEfweF+eYs4hnm8hci6Bm9Kd81ds4qHlOv0ebMhaJgoBcVvztU0DdEaqWKGYWmpzYaFB5dQcuyZHmF1uUW93OT45wGh/CdvssNiISLQFNDOx4RulArQSlJKsBD4f3z5OGBsaYchys0u7GyIQSGkIAoebszmssXQCn6MvPsjc9EWWTz9N0MuUtcawuk9Ym+e5cyqbyfYa76kQIXDdgDiMUMqSSIlVkk4cEccJiXYRrmJ8pJ/Dx84xXwtpR5asAkdaso5DM5aYKKE7t0i3VsMfGkp3PAGrp3r1hFU6gasTnG6LWstulEBOFahk+7nQOsOplSepBwt0TJOnlp+gHXapBgpPpG7LxWiefz/9eRwh0TY9N6Xs9PGK7W/CJ0BKiUuAEj6xadKKWtgAhLFsKe3EjFvuu3iIerTImfZZbCs9Wk4i6fdGuGniFeRVGRCpYWoUWmi6q8dqAFjBjspeppYvMNU6TY4COrHknRLD/mSaGKccfNen5FQYL2xmpDSOK1ysXZU6q4bqasLZparcro3VpZHJtfRpkTonVup1jNYUMoq+YoG4a9kyUUGHHcYHC4RxzGKtQyXvELYFfZ5gJFDMXVyg1uyQCXwi6zCz1GW21kFby8TIEI1OSCeOyXiKiaEqtUaHi0tN6l1NJ9YpRmKNkorAc8gGDq6b2n3ZTIZ8PkM7mqMZgiss+axLOR8QBH7vYKzUWF1NNktMmratXLc3LhZpn+92wd4xFKtbsQwgRMqjrDBIKdNgQNQljCK0zZJYxfholXseOsVyO6bVivBdD6kkpYxHs51gIkPc7lKbmWVocARrQUPKyQAhLtlqJ8Q3G2y9SXWEQyUzwNn2KY7M3U8aM3LodrrI2GPXxH7yTp6iW2QxnGOxs5gCVUgsliQWhHEH38tgrcUVLnmnRCtqEUZxutgtCCvYXt1Nf2aYpxaPMdU8TZh08ByfDcUt7Bq4kpJTJkkSNJqK18dLJ1+HkoqCU0GLXszQgi8z3LLxNUQ6TUHw3QDhVXntjrezusFFrboXe2nP6Zgo1s5gELC28dxa/i9r7x0n13XdeX7vuy9Wrs4AGpEEQAKkmEVROVCMkkw5ybJnR7aCJctjW+PxZ8f2Z7yTPDOedZz5jG2tbVkcyZK1tka2ZFFWsBgUKDHngECAAAF07qqu8PK7d/+4r6obkkzumlv4AN2orq564dxzz/md3/mdvNTNKQrj9fIip8hzsjwnTmOyNCVJUrIiY219Fc/1iZMIw+OnpOVqahXXNMULB8+NsVDMTzSoAidPd9gYRNRrAf1YMYxSHFvi2oLdM1OgjZ6mKwXVIODoC8ssd2O6UU6YZYY/nxp1OFsqBkmCG1k0axUqvkOWpdRsm8t3Ndg1s50L5upMNWw6MRzLXbKiKDnzOUopZMmo1ONdbxO9eVk0BLssRm1tZJBSYlkStCH8KAU6y4njhLSAVDvMzrSxUAwHKZ1eRLNukouKoxh2NshcjywNWXn+JHOXX2Ho6ebIQRiSmyhXrhYjr1UuipEHK+//lD9LVTQRElzpYgnJjF/j0PxlXNA+gESyt34YVzRwbA9pSTzbQ2ITyIphR5Ye1MHlDXtuIVEJdafBuLxdGtlkMMl1O15PUbyaQhdYUmJLp/S4ht+jAZ+Ag5MXmzWL4S+Nqoa6lLeWhWEIJsMN8twYRVEosjQnzzPiJDYShqogjSPyopRJjxKyPEegsXSOXWr+a8shK8xFzPIMy5I4jmn48BwPabu4gc+1r7yO++57mP4LPXAdhNAISyBt2L9vBteRhLFmrdMj8Bwcqemv9+iHKY7rESdGP9JxHOI0pVlxcKTpne0PIzzP49TSOsvdkJVeTFaq3ildILTGEhau64Iw9OD1jQFx5lH1HQ7smOK2V+9hW8vBQpPnBY8+v4EsawTmfYzhC8TYyEc2Oi4cvqwmEhjTEM77JcchLYyybJblCDs3IlB5TlrY1KoVGvWAYZSzsjZgftYHCiqOIBrm9Gtt7MBn6bmTXKqUkZPbWjrRGA+P8bKqXBGmd0OMX4KCC5uH2F7bi2M5eI5DmqRIYUKEsvbExTOXcPHMYcCi0GrcYTXqllLFSDJaUbGqeASorGAQ9822qjSqMLLQeekVkyQhy7PxdqtzRZEnFGWTRVE2XAjNmN9tYZAwgcCSNlrIMX9E2jaO7WB7HpZ0qVZc6lXzujQvkLZNxfeN5r5tm0KwhXFAYiTHJ4zYFCXNACMBbiFMZR3o9vq4jo3n2eyYa9Gs2kzWHKYmGtiW0bePk4Qiz7FyRTRMGEaZuVZaIYXAdlzTAJ4rHCnJlGBpdYMwTIjTLmGWEeUFcZoZ6ZXyeIoCsjSj6np4gUNiWSTAMEqJkozvPHmGi3fW2NacRQiF48AwSXCbwdi4VenxhdyUIPz/onnzkkavlBrtsCZZKN8wzVLyLAdh4fpGxjqLjZyyUjaeEzA31WZhfUhnkOF5ZtpGqymY3b2LvnLYduE+NlZWyYcRsl4FocthGWL8eZt0YmVUrUr5h5GGjNbKlPG1Z3TLw4IsTMmtjEiHRtVWF2hlvJ/GhGxFnuPaDkmajLFtyg6jUUgDxksX5XO2MAlvr9/D933a7XYJ5ZpqoMRCCFXKA1pYltHZkeVS1iNj3OI/DHGvtAg2vZYl5LiHFjCDEDKjhOZ5XvlyNQ47x6QrYb5qpRHaxMqbn1fK+VkjRYgC37ao+TauI4ijENc1rYbSgprrEg8ThsOQNM/RSuFIG2FJ+sOExfUNFIJaNWD57AJxUu5USpmKtQav1Jj3HMcsptQYdy+MqaiCet2n6nskWUqSZPSjjNOLA/Sl20rNUUGSFdjOFjMt7785I+MQBSW/vsz3XpYagrAlEuNNrPKNHMehGrh01wuyvCBNU0CRp6nZwrU52J3bJjl6eoXVTkh/GBE4BVrb1Gttcm8ep+LgbVtltbvKbL36PXG7Ht90jUZsublZUZg2vPK5EdJT5IaQ5lcrY2nxseKXKA3CMgYuhVWqdZmLqIQy3ncUJ4/5KIJR/+p6p8vtn/wMDzz8OK5t8WM//DZuvukGrFKgSesydBlLbIutp/N99ZKtfblb+SiWEGjMUIaRB7Msgec65HlBlmUmxNxyc01V1QJMsistG9t2yPJ07OE12iBeFtRrDp4rsSyoBQ5Q4HkC1zGiWIEH7q45BsOYTm9InimqnkuUZvQGfTr9mEEY4Xgugygrm8RtsLRRgxAC13FwHBvfc2hVfWzMJJIwyVkbDOiHIXm/YKJRoeK5JHGKbbscO9Pj2VM9ds/6VAOXMBNI20WUMTtKkGRZ6aQsRlNWSrfxkgzLlzb6MnQeceKlNBo4qjAfalmj7TonLxKyNCVVPpGymJtrk2YZ/b5ifT1hui1MxTjr8sIqDKYm2T1Z5fGnH+KtO3eWvZyCLEt58vHHiaOQJEsZDEPiODFkMa1J04RDFx/mVa+6lk2T11u+litebzH48sYXpSwqlB5cU86GstBYFKOkcZRjlF4kDCP+8I//nHu+fT+WlBRK8ae3f4bDhw/j+x4vnH6BvXv3MDkxsRmabfE8m2Y3Ot4yES0NO89ynnn2CMeeO85NN1xPEGxyR8YIDBrblhRjwps28XhZjDG+xpxvXmQUucZxPbKiMErS5s1wpMUrDswx3FiiUBl+xQJVICW4rjC6RoVxZivrfcIooyg0cWL0ZyzLJs81nutSqwRUPA8pJUmWI4SF0Brfdaj5Lp4jqAQ+9aBCNaiY0EsIVnp9Ti4uc2Z1nTCMaNZr2FKSF4qTyxG/87mn0VZOs+4jnICbLgoY3TpdFKVMd0GWD0dJk4G2y7xPvhxPP2rcEKWE86Z3MhWyUTlZlQL5aZJSKEGqbFqtBr7rkGSwMciZbLk4lmai5XByYIYYnFwYEKiUPCtwXBeNmaxx730P87Wv3WkkJJQutzCBJYxBXnvNOa65+mqktEuj2ozzy+hovDNtJaKIrf8tk+Fv3ftdnjlylDe+8Q3s27dnLBA6omdprfn8HV/hnnsfQDouH/7g+3jsiSe4+65vsLC0ytNPP81nP/cFbnv7zfzs+94zzshHXvjsuQUC32eibITRWw5mZNALS0v85m/9Plme88pXXst8UB3vMJuEMGPk9ugcROnZR7lOGeJoMFKMWpPmuVF/0CPhKo1tWXhSM9f2cD0Lz7OQlmtmb5VIXRzmrK0N6fUTstwssGGUEjgu1cCl4noEvke71aAZ+ASOa9SZpYW0JZ7jEjg2jm1hWeC6ngnL0FSDgDBJONiZ59TSKk8dP8VKp0OjGtDphYRJRgb0koiFfsrcrI8lrTKCwDTbqIwg8LBsU8stsgwhpUGx0OO85p9m9CNPWX4/8lCFpZG2Oyb3FLkGnZLEiZlWYlv4vkerEdAfhHQ6A/K5Jo4jCZycIgtJogqWJ4jzlDAKabouAoFjO9z2jndwzzfvpdvrAwJLGy11JcwgrsWVNYbDmEajVh6aQKmCXq9Hs9kyUKpWmyj2OG62EeSMlH418LU7v8G37r2fbdt3cMG+veXF1SU8K3j+1At87vNfJslybrvpBt70xut46OEHERjuiBdUiNOCs4vLxtOM8HNLcOTIMf7Tb/0eczMz/Pt/96v4vj9WFDS1CXNNbcclTjOy1DRGjwVZx9s2m/kNm1DXJqC5+XWzfc5CFWocGkKJvHkOlarNBbsauI5VMqPNdUVDGOWcO7VBOEiRloV0NI5t49sOrXqNqYkmtUqVeqXKTLtFM6hQDXwqFRfbdpDCUAJUibolSUwcm9ypVq1gS8mErjIz2WR+ts1Ms849Dz1GZzikVvXohxESqHs+kSqwXdeoceSKYqx0UGL1RRkGAirPxnmnKl4OTm9tyYiVkXgQwiRKGjP0TJXzhVRRkKcZeVZQeB627TE3PcHzL6yythGTqzZKQ83VBLIgSUMSBbad0eks02y2GXVkbdu2jZ//0If4w49+lH5/yAfe914u3LcPaVsEvk+lWiGo+FtCCLj3u9/hz//8dt552zt52623nIfyjMILgdrEuIU57tm5WRSa5eX1LXnF5nnf8aWvsryyRrtZ57Z33EQcx5w4cQJZDnFrt9sGFeluoJTCLlEFjSao1Ni2fRenTp6g1xvg+X6ZmIvxUaE1juviOh5JnJKmSelnRsfA5j2wLLBkKbVoGj4sIceh38gxjfYTU0cpkFKaMFQIUlUgLYGUm/2kozhYaMHqYo8sLWhWAwLPJ/A9k5BKm4mJBu1mg6nWBFMTbWYmp6n4Po4tATP/SZNvWYKU4d/XPQAAIABJREFU41AzipFxFiYB94BqzafdbBJ4Ll/57gN4Xm7uERrHkdQRNIIAS2sKNKoozMSWUT4nR5KHI01+s/tZ8mV4+lEf4qgJwbFtLFuiC4WgQpHGprycm7gvSVPywkj/IS3mt7V5+ugZ1noZYSyoVQQVF2pWxuq5JSgK/PkWy8vn2Lvn4vLQNVJKXvfa6/jiHV/ksSeeptWq84rLLgIMfdUqxzWOLqzWpnl9eaXLPd/4NjfecAOu65x/MqMagNbA5sS9Hdu3IbRgZXkVrc2io1xIq+trfPe+h1BKccXlr2DH9lmeeOopllZWqVVq1Bt1Wu0JhGXR6w1JkwynXIwg2Dm/nV/+yIdZWlyiPTXJaHZIlmX0+n1816VaqeC4Hq7vEq+mbPT6IARZmhLHMfV6vSyva44ePc7Zs+doNpscPnwI3/fJipxHH36YYRgyM7eNu+6+m8lWk9tuuw3PDyjKBZOmqakLhGHJP3fQKFzPtA0WuWZpocf6ypDpVoupdptWLWCi0cD3AlxL0mw0aLRaTLbbVCo1KkEF4xcVurymRQllb5XZy/OcTBl833Q3mVkHaBtpO1xz2SGWO+s8evQEacUjy3M81yVJcxqVoPRco3xr1M3HWNtyRLPQQpfI0csoTgHlxAxjXplWkBblKrMRlo2l8tJDF0RpRJLlVAootM3czASFKtjoZwyjgkJJHKFwij4rC+tUgzrRMGB5caFEY6zxFi6lzZ49e3nsiac5feoFLGEDRbkAS1xfUBZ94MrLL+eXP/JLBp3Z0pCSltXISiUwz5fJzmDYp7fRp1qtgrBYXlktd62yCissjh49zvJqByHguldejbQcdu/cw803vJVBf0CzUafdbOHYNlEYcm5xhfsfeoTFxbPcfOP1TExM8vGP/wW5ytk+/z7a7RYLC8t89E9v58ixY/ieyy03Xs8tN9+E7/nkWtEbDCkK+Ls7vsKpk8/zC//i57Ck5Ctf/Tp/+mefYFga7c03vZX3vfdn+NRffobPf/4OJicnCeOYldU1PFfSmpjixhtvGBueKVZZKNfHlmZgWbPdYmp6koWzZ5GWYLAx5MJd29m3YwfTzRYz000mJyZxpIshfkik7VKtVvE8DylH0ucWShVm55HuOO+zpGUcpGUhhTZ1nDTFyJCXkLhWeJ7k2lccptvtYa2s0wlDbNvBd322z81RlBNPRuG21mbIm7EXRWGBVgVoo+3/8iDL7ylMGbjPTI0bqVMVRW7mEmmLIjNqxpnWpNi0mg0CzyVMcjY2IvLpGtotmG7bFColKXKGw4j1tUXyPMFxAiiNWQjYt28vwpI89sQTPHD/feSq4KKLDrC0cI5vfes77N23j2/eey+2sPjZD7yfo0eO0Ov1uewVlzA1NcmzR47w57d/guWlFbZv38Z73vNTHNx/IY8//jT/15/dztmzS7ieS17krK6tkSQJvu+Nz/e546fIsoKgErB//4WApt1q8uEPvp80TfBch3qjhud7dDb6/O5//yhHjz2H0gUL51b4xV/4OR546DHiJOYn3/0TVKs1/uxj/5Nv3Xsfr7ruVRw9cpRPffqvufSSS6nVa2ilCcOY7zzwKB//i79GCsXb334rjWabT//l5+j3I378XT/GPd/4Bl+/6xtcf/31PH/yDBs9MzCacoxUnCmePXqE69/y5tJxbToB27YRlskPup0NhDDFq343ZM/0LK+58iqqQUCzVmVyagLP9csCnEJYpr/XcbyyTrJ1XCYGDish0yJXCG1hydJWlEZbGmXbY+7MaEfQWjHZrvOKgxcgLUltMKAziMiVIokj02E3Ct4ZhaxqEwEbxfJlEfClPP1LNIaPPKooL95oBqnZWqwSPQFz4kWeE0cpqhDkysJ1HaYmavTDhKVuRJabeHqq5SMt44XDKGXY7xKFg/IzNz9/9+6duI7k6WeP8H/8h9/iN3/z/+SZp49y4sRpPvf5L/M//vjjPHv0JN+8937+9vNf5PiJU3z5a3dy8vnTZFnGpz/z1zz+5LPsu+AgTzzxLB/72CfY6A24/ZOf4Zkjx3nda17H/gsOoJSi2+0wGA43YUatOXt2Aa1hamKSyYn2OL424YEkL1JqtQrVSoUwTlheWuXmm29C5YrTp8+YwpnvE8cZ/cGQ48dP8N37HqLZqPP+n/7fOLD/AoZRRL83oBZUUQU8/tSz/NGffIxuv88NN97I7PZ5vn3v/Zw+t8Dsthlu++G3c+jwIcIwpj8Ycs0114CGwHf54Pvez/4L96MVDIcRZmToFt2h0S6pyz5WLRn0IuJhAmHB6665hh2zM7Sbder1GkEQ4LqOmQTveXiei7StcWiodEExmmhuWSDsElEzE2yyNCcMQ8IwZBCF9AYDBoMBcVm912yluih2zE6xY6rFTLNKo+IZQdhwYOoQowKeyhnxVsqoZxOsYHPK/T/Z6PPMaNKPko8sy4gT483TODHwVlJyQYRAFTlxZJTECg0Ih20zk4RRxMYgJUoNxloLbOoVidKKfj8kjAvWO6vliZm/WsPc7Ay1WhWlNGmekeUKz6+yY/cusAWu7/KvPvIRapUqx4+fZHZuhkznnD53lhfOnuXJp55lZnqGD3/4Z7nmmis5duwEz514nudPnaZWrfETP/5D/PIv/izzO7YzDGN6G73yzE3D8Xq3i9Ka6anJclqhOTbLsnjyyWf47F/9LZ7j0KjVKQrNa197Lbfe/BY832FQ9rbWKlWyPGdjo8+zR44TRhHz27Yx2W5R9X2TJPo+1WoVheLrd97NyZOnuPrqK/mxH/0Rjh4/wZ99/BOkeY4XBDzyyKOcPHGSShAwNTVFEBhEaM+e3bzztrezd89uKBR5VpxXnRx/Lb/XRRlb5xoVFuzftYcds9PY0jKcHWdzrM3onEfXxoQvRoSwQIFlYO28yOh2OiwvnmN9bZlUFdiVOkGrRaVZo9FuUKvXqFWreK5X5mcmnMzzDMeVzExOUJUOE4FPRQpUHJPFsdEd0hpVpGidG4hVmhBK2jaO6+A4zjhpf7HHi4/fseW4IjjCecF4OiksVO6R5xlFlhiGYVGQZTFFlqGVjRKGZnz/Y8fY6MdmMl8DbKm5YGeTJ05ExtvHGUsLZ9iz51A54sbcoEajxcz0LMvLXd72trdz+SWH2Lt3DwUFlUqFeqXK4YsO8hv/5teYnp7m8aceB+DUqdNEg5D1ThfbtnnssUfZ2NggyzLCMCJJUqanppiYaOO7Dgf2XcC5s8ucXVzhggsvQGtDEIuiCK0V7VZrjHQIYYzg0cef5s6v380b3/IWJiYmEMDU5BStRoMg8BkMhoRxzNTUNMWzR1lb67C4tExRFExOTBD4Ph/+4Ae47Yfexo4dO6g3aqA1aWpCBd/z8T2f546fZBCGABw5/hz/5b/8Dr7r8u53vYv57Ts59fwLpo5imQHEruuY4k1RlGnzedUJtLBQCPJC4VgWRVzgxJJ9u/aYWkiJtVuWRZEXoM2E8yJPy8qrgyVHREAzHTHPUzq9PkJa+NU6frWFyjKSKGTQ7RKHkRmMPGLHOraZEl2ig1mWjSvs7YkmlSWPKE+QEtLhgH5nFccLyLNNYMWcjIGlzSy70cTBFw9tXtLoLWtz61AlUUtrXU5qHklKW+PYXylFmgyJ4iH1mo+2LaYmGwgh6A0KNgY5021JrSppVAT9Xg/XrdEfxDzzzBGuvvZ6rHKVCg2u47J7106eePJZpqYmeOObX4cQkBc5M5PTrKyskuUZ11xzFcKCjUEXx3E4ffoFLjpwENu2WV5e4Xd+9w9I0pRXXn018/PzJXJRjq2UFtMzU+QqZ2FxaRQwYgmJ7XgoNNWakeygxM6zLOPZI8c4t7TI/fffz9TkJAiI4oiKX6FaqbKx0WM4DGm0myhVsNHtUJTj7v3AByGo1WtYyxZpmlFvNNBofM9FqZz7v3sfDz74IG9/+60cPf4cX7jjS1x+6SW89U1vZO+ePczOzlIUOZ5rKqIjw5HWqGBj+ETj0nK5ACwE0jLzmywskl7C9qkdNGqlFhCCIldkaWaSeiHBcrAcYXhEloXSOWkcG/pyliNsSaXeoMhzNlaWiAZDpO3g+QGu5+JO+mCZ4pcZ5WkKXiOtmjiOCcPQEPd0Qb0e0Bn0EJYmz1P63RVa09vHO49lGW8+juFR5HkxXtovL5HdMoRsVOEcNyiUibtt26hicxvN85wkjsjyjNS2qFarNKtVBmHC+saQdJtEaItW0ycc9omTgrUn1zh9aoUf+pH30Go1GRVghBDs27sHKS1Onjg5/mzf89m1czsnT55kaWmFyck2KMHczDYCv8riwhJXXXUlu/fsYnV1nV//1f8dKWHf7t14vs/k5ASdTocoigkCj0ajiUCwcPbcJunLspienDBJ1JgHZC7GwuISx088R1oolpdXmJpsI4Rmo7uBdDzqjSbF2XN0Ohu0202woLPRo1KrQpn4IQS9fp/bb/8Er3/966jXjXO46KKD7JzfwZfu+BKf/OSnOXTxIS45dJgvffkrbJ+b49ZbbkZr+MpXv0qj3jSjMYUoGzQyHMdFC4s8V4wGx5n7Zm6lFgK75O7YyiIvBPPbt5uFogzSYtumAmpJG8cLSFPjtZMoBgTSMVCj61fwqpLhIOTsqVNQFLSnppm6YM6oPUQJWZaRJokZ2ZQlaK2QspyfW5psnCYlsmPQHcsSJRvWhMz9Xt+QC7Eo8twM9S6ZAiPWwMhZvmxqsdi8ZlueE+e9QEibETNBYDQa8yRB5QUqN8nu1HSDheUeK52ILK2htGC6HVCvunSHGcMwZDiMWFhYot1sjr2TBnbt2oUjJefOLZBlOb7vIoRg7769fPXOOzl95hSHDx8EoNlsMDU5wdLiMlmacuklh7njS1+mVqtwwb5dPPzgI1x19dVcdHA/X/nqP/DUU0+z/8B+7nvwYTRw7twCRZ4jbQkCDhw8yFf/4S66Gxvnbat33n03nU4XaVkc3H+QYWgGSqyvmxxgoj2BUorV9XVazRYCwcrqCq+96CAWkudPnuaRRx7jji9/hfsffpRcad76+luZTPczdfw1XB7fijp9I+KM4O5/3sJqvYZLui+w8jW4a99jrKQn+du/+xxvu+VWLr/iKixpkeYpSmn8oAII0jzB1CM2BU1HTt/zHGwp0bGFa3vU6lWSPEMB0naR0sWSNuEwotfpGa9dqVBttSnSlDQ2yXeamt/xAp/p2Vkc2yXLMzora4TDIVE4RCtDMQkaNSYak0hLkuUJquQEaa1RWAzjFE1RMlM1EoErJYXS9Ls9hmGIkB5ZlhtuzRbDHvUo6P8/jL7YIrO9law/QgG0Vpu8nEKXUKOmyLNyuniCKzPm5yaxhEUWF6SpRZZpAt9hou4xiGKqgUcYRpw4doSLD+5HIEjyBKU0MzPTtFoNVlaW6G308LwplFLs27sXx5Y8d+K58TF5nsfMzAynT73AwrlFbrrxRu686x7+62//DpMTLY4fP8573/PT3PCWN3P3Pd/kD/7HH1OpVDhz5hxaKZZXVoizjIptCMGvvOpKms0Gz508wWAwpNFocG5xia987W6yXLFrfgeXHD7EkeMnkVKysdFBWpqd89tBQ7ezzoELDwDQ7XS56vLLec211/Kd+x/k3/zb/0QeK1rhAXY89OP073ktt524EVv5rN4n2IXxvtED5vq/ml9HvVDwwi9FrAXTbJ/vctX+1yOcBMe2zdRxpUzYUlbIdalWoMbQYAn2qQLXselnGTW/AkKQJBlaCbJ0wIbu49o2jVYby/GIhkO6ax1GnCI3qOIGVYK6meqilWK40WdjvUMWJziOg1sJaLRaBFUfpSGLE/qr6yZ2dyRuEJjYXguETk1l2bZxhKBSqVH1fDasAb7rESYxcRzjVWzyPDWNJHk+7m3OlaGlF2rE+n3xx0vi9KM33mr4BhUw0JXt2+gsQgjwHAFK41mKtL9BLjWZVTBZrSGmQOiU/tAMR0iznH07ZqlYQxzXIYoTjj/6bR6drjM3v4tHH7qPqh8wuW0X7Yk2x449x+LiEtPT0yil2bF9B1OtNs+ffJ6iUKZZQVpceMFennz8SVZWVrniisv5xZ//ef7kY3/KE089w875eebnd3LJJYf50Ac+wKc+/Zd01zpcddnlrCwvUfEc4igiCHwAdszv4K1vvZ7/9Td/w6f/789yw/Vv5VN/9decW1jGsixuuekGms0GM9OTXHPl5QRBQF7kvPbV13HPPd/g7NkzvPY1r+HwoYvZvXsn1WrAr/zKL/GZT3+OR/5mjX3Hf4zp6BCOqhIicP7RO2EeFhIrrzHXv4rZZ6/gyM9rdv5Yj3/9wd9AV/vUahVuvvmtHDp0EY1mHek4mxRtrUbQzZianKYFOJaBD5XpynIcG9dxkVh0l1exPQ/X96lP1rClRZYkxGFCp7eKKoqSCg2259KemMBxHJRWRhlaQzSMEZbE8wLqzQmElERxSJqmRGGMKgqyIieMYqRtFkMDgbe4giMtqp5L2A9RWYooyl4I1zNU5pKjVOSm3yJTOb7vv6SnFy+W7f7LX36/Hhn99wmAMpqirSmiARQ5tmWhMiPoWQlM+diSJW1HKfr9PkVhuBlZbuYajWBRKW3iOKZaDWg026x0NqjW6rzq9W/mwceP8PW7v8Ev/eK/4G233lqGXZqlhWUWFhe46srLyzZGwXAY0u32qFYrNBoNlIbVtTXW11eZmZmhXq3iODYgWF/rEEYRU5NTJIlZuNVaFSmlwX+FoNfr89u/99+474H78XyfcGgSrute9Up+7Vd+mVqtWrJPTaw+6tvslKq+7XabNE2RUuA4LmnX4uk/CnjmT1yK0N5KpkBY4E0rKjOa1qF8HFrGK4KNIzbxikWRUnJ3GP/ezKszrvy3fSYvN9d2VFGmBB7GGDegi4S7v/CHLD7/NAsne7SrdV5x0YUEnovv+xRZjmVZTE1NYWEx6PcZ9PoU5SQ/KSRBUKFWrxFUqnhepczjYnKVGnp5nGC5Do1mE8fzEVIaBbcso9frsbq2wsaghy7zBm2ZmLxaqdFsmoT+2aePcOL551npDziz1uXAVddRaU2xsnSqpIObewgGBi+0QqFK+nnKx2//wj+azb444Ux/Pyl/cwGUrL+iQChFHiUkeY5bDupShSCNMywJ1WqVLCtIYkWej3BeSRRlFEoRRgm9fojj2rQKwfLqaRzPBSF47sjj7Jhq88orDpdTq01jt2VZ7Ny5g507d5jFJwy2HwSV8/joloDZ6SlmpkYy4iMvoJmYbDGBmUjo+fUfcI6aRqPGv/6Vf8lXv/p17r7nG/T6PS65+GLe89P/jEazwdbHJqwpDIxZ4uS+b3aO7jM2j/zHKue+7phtHUBoKnOKuTdk7P3RhMaBDKuRUXghUR6SFxk6F4jEJT/VoP9InTNfrLD6oIPKzH1Yvtflrp9qcclHhhx8b2QqrhpGnVnj0KbE3F3XxbZdU0BSmt4wwnMdBv0BjUYDrRQLZ87i2BZBpc7E1DS+5yItQytIktjUatIc24mQlqmyCseiUm/RmHCxLJv1tTVOPHWM9W6PTr9nQhspGERDkiRBqQLfd5mYaBuHVGtiS5eg4rNtbo5zSwt4SYTvSJIwxK+bPEAKy0DHupwVNuqfQYwT2xd7vGR4M/p6Xj+iedLMCQ1D8mGELSVB4JFnOf1+jOMYtYQ8zXDcALCo1xsMwyFxnhlKstbkhSLXmiTPsT1JmucUWuHbppd0eWGZw5MzHL5wF/v37UZlObbnboWef+Axn2+8W5//wT/f+tgssRvDr1YrvPO2t3PLLTeR5zm+55hk9x/9fPN3BPEKIVh/QvLND9TpPzdS9dUEs4oDPxNzwbsT3NmM9XSJpztHWDr3AomKsfCQQoIoyFUONY33Rp+pN+9i6pHDLH1iF+sP+uhckKxKHvmPdSgsDr4/QjqjRpbvL8srZSC+rCgIo4R+GFOvVAg8l263SxAEVJsGTZLShCvdTgdbmmPPC4VbQpLSdrCExK9Ux7h7rztkcWGJFxYWef7cOTqDoSlilUTIiu/RrDVwPdeM1EwLut0+k81Jw7JUinqtiiUErrSouzaztQqpMhLxwnLHyI8WMJqx8f8CogdeauZUarQWHccpb+jojY0QU9zboAhDPMehWq2hlCKKEqOMkBfYrsNoW7WlxPNcusM+eaHI8hxHSqP+KyymJlo0mzV0oRhGUfleMYPhkO/c+x181yZwLK578y1Mzm4nCIIt3I3xJeA8uOk8oMkYmlKbVIfvC9nEpqGMK8OIMprQRg+zVBD4QY/xtRHnf0bvuOSb72/QP7E5QGD6mpyrfnPI5JUpa/Ei95+9n064xlx9L1fseAtVZ4LAaWCGJOfkRUaYDdiIVzjdfZalg5+h+Z9n2P61N7Dw+wdRsUQlgkf+cxXLgQM/E5oC0padeuS8jEqCTZxkZCgGw4SkkVOvBnhejTzPiZPIiPc6gkpQozXXxq8EqEKh8wIr1+isIM5SzBw6hRaaOIrprvVY7/bp9o1Y2Mxkm6BiKA21SkC9UqFRreN7HloohmnE2kYXVeTGi2vwXId2s0mn22W6VuPaue2cimPO6ZL2wCYFYdRGJNCG9v4SUt0vUZyyxhdqJImM1mgFWW4w4ZltbcLBEN/zGAyHpktegFCKLEnNOPvcjDMfRCFJalrZBALpOORK02w3kNIiSTM63R62tChUges45HnOMEpwpc3jDz2A79hc8errmdu5G8fzvt94z7PjLTj1CGW3zKQ983ts8YZbvgdjbJsvGhvz1t8bGfBWgx99P3pt2hM8/B8qY4O3HM2OG1Ou/e0Qeyrh6fVHOLryKHsnLuHqHbdScRrcdeedHDv6D7zr3T9FtVZHIImGEf/zzz/FwQMHeOOb30ZY9Hmu8wjPv+l/IYZXYX3yTaiOj0oET/x+QOtwyux1ahzXb0qnlHNobY2iIEsyhuGQLG9T9tkb9KRSJ/BrSGkRxUPW19axuzYVL8ASAk+aZm+nVA2WlqQocpJhRBwOUblCWi6t9iS2lNi2pFqtUPddAtem4khc20ZLge04CC1wbQfLshEY7Z9mvY4tLLZNTzJYXqAhBFUhiZMYhEQXRg6kKIw+fZEX48ruP9nozQ3U4wlu4xCnxEWbtRpSGG7zaMtUGrKiIEkjXOmiUdSqFTr9LkoJ4jjFtm183ydJU4ZRhK8VjuOyuNQhihMmJxrkSmFpjSqFmWzXpsg1zz76IFXfoxIEtLftOD8Es8S4H3R84uPzH5WuGVtkWYbCQHF683XlAhht14bNp7eEVCam3NxO9djIx0Wg0sCOftznzJddRgvkop+NuezXIrSX8vjSt1nsn+FVu95GtAJHnnyOK6+4gkGvSxQNS30h837ra+ucOXOKN7/lesDivm88yMzsLK+78Ed49Ie/zmDvIvzGT8LAJ16xePjf1Xnjp3oEU+eHeApT1HEc81yaFeRZRpJl9PoDJifaOLZDnimWNpZYWFzk7NIKUZhQr9SYbLRpNaq0GzWa1boR3TEXjCTNyJMY25KlzYBQmkE0IE0TLEvT9Bwq0sF3XLyKR6XVwPKNrInreqUkiumgq9dquI7L7NQUnWOnqbYbTFerrGhTeBNao7VrEtk8Lzn63x/Ofe/jJWP6743nKW2m7rvUXZd+r4dtWwzjiI1hOBb73xhEOHbGlFOnPwyJyql0QljEaUqc5URxguc4gMXyaod+GOG4jjHeHFRh0s4oSWg3aiR5TpxKnnn8QbIi5bW3/ASN9qTpDrINZiy+5/h/sCZKeU7GpUOZ8G2Gb+XvshkWGc+/9WKWiMgIKeH7c4TVh22e+aOAUadU+9KCQ78QI7yCx1ceYHG4yGv23sZgNeXPPvrfmGi3KZKEQxdfwgMPPMTZM2fYd8GFCCFYPHeGvbv3smPbPOfOnuHLX/4iQbXBz33ow7xu34/yWOVuTt52F/anbkJowdqjNs9+NOCKX4+wbIySgKBMcgVpkuH7LmE/JS9y0iyjMxziV6ucWVhiY2PA2YVV1jd6DOMEKSVV16ddW2Oi0WDnzDRqMsMWAlfaYBmVgihJ2BhGDOKYKM+I44g0SfE8j3qtju+5ZGlKt7uBP3DwopBqo057so1lVbGlhS1ttK1oNVvUq1UypahPTqLDATumJlnZ6JFrSkqx0fZR2nxP2ajyYo+XnDlloECr/CuxlIY4phUEZGlCXuTEaUoYRkRJapAbLRiGCXmh0ELQ6fUJ45RcwTBK6EcJG2FoxpkLwTCKGYQxCF3KVY4MyujcZHlGlCYUQJIX9AcJzzzyMPfdeQfhoF8asTW22HFQo0exnx4b5JgqXeLXW891tBg26dTnJ7+jn21dQOOG+e+pZaDgmT8OSNbN671JzSv/a4g3qXihf5zTnZNct+sWmu4sc7Pb2Lt3LxcfuoS77rqLqZk55rfvYOHsmfHnT03NctNNb6daq3Hnnf9AkabMz83SajZJw5yLJ65j/4di1LVPmcWpBcc+6dM/bY3FpkZ9t5qCoOJhdKIsgiDA8wPiuOCpoye575FnePjJYyysrKOUCTPq1QrStklVQZTn9MKQ3qDPMIpIioxCmwb+MEtZ6nXpx4b6rHLF3MwsO3ftYmJ6Gr/ZpLl9GzsPXkQw0UTa7pivL8f1HzOw2nMcfN/H9lwOXnqYsB9SUxqptzq3LUCFsErNoZcR3hRpBrbCKsk9+TCmt9ph+/YZ+r0+cZwwTBLyLCfLDSvRsszIcoGRZUuLgiRJyfMCLEmcmTizyHNcx0Zpg9kjDORnyFMFWoFwJFmeY0mHQRhSq0ySKSPksd4Luf8736S2rc2eSy9BeiUvW4xGKI8MdPPfsfceIytbLhojTz56ffmzUVQzMpgy0hlH86NYGUoRIpNlRSd9Fr89weiV2364w8RVilCFPLn8IJdufyUnnznL3ae+w5ve+Bbe8c538t//4He59JLLqVRr/OR730Mv6/B8/xljpDvv5bs2AAAgAElEQVQqNOwJEJrLLruCIo15xzveyXNHj/DlL/8dfqXCP3v/TzP80DGWH9sPkUfaFTz3d7D7g6tG8weNUDm2D44rcD2JUhn1epVaUGVtpcPxU2dQhWJ2qk3Vc7GlbZxSFJPnRjS1F4ZYKFzbYqrVouraONUKOkuJNjYYpDm94ZAkTWg3GgxUzncef4rnTp+j0ahx9eWHGS6fo2lpZlpthF328pYkuLwwPH1p2Uy2WthCUG/VaUxOkGONd2BhjZgA5d3SGqFf1I+/tNG7rlvKXwuyOKFIE6amGnieZKPTJVOKKMlKFMVwo9Msw0LjuY5ZwWXhIC+04W+rYpw0SmvUZK6wLG0qgZYw/P1CoWzIc9NbGSWaMEoJPIukKHA8h+Gwx1e/9tfMynuhLcshaqZTyGDUm57cwjovODG2uQW+HP8jGEc0jAxcbPk5W/KC8dIov6px4Uh+5k3Yq/MAqFpI98a/JxNv5HT3GIFTY1fzIEPnKF//+lf55jfu4e1vewcf+Ve/SlD3eGL9Xp5eu49OvEShM1R5ZjWrwd7aYQ5f9Gr27f5JtNb8/R1f5LYfeRef/8JnGayHXPnmi/nWdSsM79wBWvD4X/X57us/ivJMed4tNJPDkEBAs13h7KlVGvXGuJawfXaa+dlZfFsSRymdToc4Ssr7IIjCEN/xDKtRayr1KnPNGn6liqsUXqePdFykYxuV6qKgPr2dY/c/RaIE8XoPuz7BRdvnEKuLyCLHEpQda6aPQeucLInRRUqrVmXx1FkWq2eZ2bmdE70hhTayfpbaFPACQ4pUQqPVy0xkldZYCvIkochTcinY2DDbWpYX5GW1VekUrQ03Os9zHEeSpAXhMKbVbrC6ss4oZNEabFuWE8lN25cqNNKzcBxJmqRkWUql5iELIwCVaUG338O2G2yEIc1KgO9XSNYSVo+uUr+qCbI0QkUJYqkyvNlq3FvJV4IRWU5tgbk243Ij0DT+fS1K+Y5N4x6FUQhKkVKBiD3aX7pivBzyy44ydcBDozi1foxLdrwKW3gcPnQpb3rzW3jy8ce54IILsRvw9VOf5tjGY+SozWPTBYWCVMV0Oys8132Ca7fdwr7aJdTqFT772U/xisuuYH5+JwjNhe9KeOweDYVAHN9O8eh20qufMKX6HJrKDGqenKxQq3koZQp+vuvRqjdxpE1vo8/66jqO67Jn524KDVGas9rt0uv3KHRBmDnERYEbBNTrdQqlaLSabMth27btrC4ukUkb4fukUUhnfY12vcZU4LGz3cSruljCQIz1Ro1aEKCVMvc7TYxQq4BwMGR9ZY2Z6VmKfmQoC56DKpNXQ0ku7VWpcTj3TzJ6rY2kXhxGkBp+RlKOvlld28BxPaQtSZJkLNtthIYsVCnvlqYpvusReEZQyJaSMIyoBw1TmFIagcHq0zTFdSskaYIQgo1en3qtji4UuSroDQpqFQ9pCeI4ZnayjSMdlr61SG4VNF7RRrqlZqPYasSj82Hc32sMVm8aLVsT3S1IjjH7MV5vdqlSDkpsQYvG4ZNALk1iLU2Ol03x+ifY1tpPP+2acTu6xZ9+7GNc9orLeP0b3sKu+X1MbJ/gzlN/yfH+o+Rk41BMjY/ErKwCRVetcM+5zxJvi/jn730/ndUOCs2nPvVJ1tdXee+7P4g3rUgWLSgk9pP7ia96YnwtbE+S5BGNRo2J6Tq9/oBd0zAzNUWn22O9u05nfZ1mq0VreopjZxd58sQ50qJgz/wc2jEKZkE1QFmCaqVKNTBNHu1mHV2KBnjSJgxDtCX4iRvfwpmzZ9m/axcXNiqoXhe76uG6HkEQUKkE+K5j2k6NcCnCEjiuw+z8doo4x5Ym3PH6PXKlkJbAwsK2/bGDSZIE5yUkQF48kVWKsNsjC4dolSGUpshy1la7xJHxxnEcG5KRMuGLsASeZ3pnXcfBlhZJFFGr+qAKPGdzyJkuCtIkRiuFbRvBJkq+tW3bpGnKcDigUQtAaZI0oT8YkmUGu+8PI8I4ompV6Ny/SvT8EIpNwxUllDb6M7L+zUS5/Ik2jW/a0qa5+XuKXIauWpSdOaqUiFZoXaC1OWY9Rg001loLEZpwQQcx8tAiTW+StXCZZjCDziR5lvN7v/87fPL2j3Po8EGOrj/A84On0NamNPl56++8GwMxQx5Y/HvW9Fkcz+UTf3E7X/n7v2N2ZpZB2mPi0tEkb4E8vW20WgGI0xzHtbGlYH7HJHE8RAuF5zkURU5vY4NCKdYGQx46cYoHTy9yfL3H890+pzoD2tvm8byA7XPbaFZr+I6DZzv4nkc9CAgciRQKTYErBPQ22Ok7vP6iC9ndDBBpjO+5RnnYltTqNapVo5/jjLVsLFzHo1qr41er1Gp1Ar9KUKmRp7lpdMkVWaZIk5QkSYjj2EClL+7oX9zos2hotG3yzBQibIfOuqm25UVOGqdG4kFpWrUKFddBaI0jJdXAwSkTlP6gj+95BK5LzfVpVCpkaTYmokVRhOe6WJYgTTM8x6bIM6RlQhtdGEloXcBGf8BgGBEnGaudPt2NIXlW4CceK3cv0n+uR9nRZ8zeMuMrNeOI5HyPLswisBCmmX9kVeWj0MogE7pcHMoYvi4UugChhHHHahMlco7sGX+YbvXxpzSOdNmIu0zUZjn69LPc/51vIoVmfn4e7eQ8u/4AuUjGu9JWNGnz/+bYRvXHSA94ePEu6pNVLj5wEVJYfPPuuzh69FkaF2zeR/vUDqzIL6uzZjZwJfARUlD1zDhPrcy8KE/aSFsSZTkL6xtM79iN1ha99TXijS4TnseU77Nr2wwV38Vz3XGhTgjTXRXYDhONJq1mE2FZZspItUq16lOrBQSVAK9awfWrVIIaQeCbHmRpGafpu/heQCWoUa81yZKEWqNBtVYr6R8jHTkbCxsLicQAGLZ9/rzcH/R40fAmjmM8xyFLNSrPGfSHxGmGY8mygGCbudVC4AjB9pk2wyhhrds3E0g8D0u7pnCQZ1Q9F0eaLSxVhdG0z8xW3h8MDGOxMEKgtrSMcKtloQQEno0qKcmDKMISkOU5YWKRZilT7SbJQsrpO55n/7sPYs95I7yyNG9ttszvqdh+7+UZJ62Uucr3uNqRYWsF1ljpdRO6BIVYa4zfuZhdI2iZc0rziMCp8urXXMuBA4d49NGHmJ2bYz1ZohMvM8qgrVHFV4yqu5s709YEXGtYjE6ykpzjFZdfzdTMHK+86hparTYn1zfdndWrYxeVcdjkOA5pP6UQku7agKQfkucpvuNS8Y3uZOAk7Jifp9WqccN11zAXBEzVAg7u3omjNb7r0qhUEUXBoNulGdRQuaJIM6QA37GY3zaNlJLuegehwLFtbMfF8Vy8oILtWlSrI2M3oWeuTTug79slI8BCShs/qOBVqgwXF8eOinEeNmL8qnGfxz/Z6Is4xa45WI7NoD8wjMWJJkIppLSNZHJJNEnSGMcOmKzXqPk+trTYGAxLREPR62xArYJTq1H1bDxtYQtlSteJEQCKwgjXMc3o0pK4jkOapqjCdN/UKwGDMCSOYwLPNXBoCkmcEHguURQjEjj1hZPs+eELkFNG4+V7jXZkpFss3YQRI3hSmKrzJqeH8yt92licFkaWfDwIga0Fr9HnYVgPWlDoAtvy0AparUne9KYb0EJzdP277Aj2ceHU5RTkLIUvcKZ3jPV4kUKo897S+DeXut2kYtdIi4RzvRNcd+BWLjpw0HCROH85W0Ji45CVS9gkxmYH8xoOzqJGZTnacahUfSYmWwSVACksdHed7UGFbVccJMtS0jRC+gGT7RatWpV+Z52oN0RPZCRxRBINEcJCpRmOY7NjskXN99jYGBjRJynxXA/XtvFds7PYoyZ0pUxbqpQGNbNAZWbii5KSBDgz6BvoQ2zKqI8ckVIa25Ivz+gdaUT6VVHQarUIhyFSKSabdXzXwXccKJO5LM/ZGIZYUlANqjhS0BtWyHNFtx9RZDlFVjDoD6hVK1Qch4bv4NqSTm/AMDJS30VuNNKllGPINC0l7iwhqHo+UlhkWYptGe8fp4rV3gDHgizKyJ4rOPOV0+y9eRd23dx8A2OOBOhgpOUuSprD2LCFoFAaoU0RZGvFVpUzoYRFmcDqLYUskMgygT5/oRXK8LwlshzFo8aFIlHCdHOVvfw/pL1ZjKbZed/3O9u7fHvtXb1Nz/R0z3A4w+EypGVKlEzapg0LlmUjMmzDRqCLGEiQOHeBLxJEDgIkdw4SGAJs2IABwzYUQ7HlOFYUyZIomhQpkiI5wxnOvvRS1bV+27ufJRfnreqREVsxp4C6qUJ1ff3Vec95zvP8/7//c5ufR6DwG471zoLXz7/Ftw5/g6U7hn5gs5te58Xtn+LG6C6ZGnJv/UPeOn4ZH9xlmRKfwQ+1a4VESx1fW6+jIsQe98bWCLHR0RYlSWLIspTdnW2apqWqo8lDhA4QKGmYTSdsbszYmOQ9i17iu8fkudBYLBH97a3EJCnbo5xhYljWDV3nYpoJnlRqEikI1mHbJlaEUqClRvRrqrOu3xhLlmXB+bqM+2wAelDUJYtT9O/1R+neKKVZLSsSo6nagmmWsTvb4Mn9HbZ3p5gsu9Q92M6xXK1YlnEBp1ozTDKa1jIdjanqmqqqokm4KvFWIVXs5SazCaukoWqiOrNpW0KIXSJjTLwUa0PXZzEppUBpbIj3jNrVrNcVSWJoGktRdxTfaDDHK168lWGkQ+kIqrrowrTO0VqP8wEX4jDMWnBIKtuhtGCQaQa5RmYZD1eCVx4UdB8a0saMqcAgT0iTOM/wBDbe/w4v8uW42x6Nee03DvjyzRaBYl0vYQBH5X1a33Bj+jSpyVnXc3xflzrrMSHnhY2fYje/yb/54B9z2jzizuST/NH9n2GkNulciw4pQzO5JAzYYFk1Z0zzTVbvJZd/x8FVi/WBQS3QQWKIsUmZMWgTOyTFuiIdj0jSjCzNybOYxVXXdV8SpWTpiGGeMR3liOAjmqSH1hJA+ECqDd4GbNUhtUKjwQgGRiJJ8EncjbWGRCuEczjbxfakkiipQQgcUW6+ripa2+FDoLSQjnfoTk4fT10vp+EX5V/cHH7kRS+CYDoeY6TgxtYmd5+4zv72BpuzMcqI/hfEp6ztukvo6HJZsloXnCzWHM6XVE2Dc0OUiN7Nqqqo2pZVUVI2DU3dErqOREhUIlBaXeY2eed6vIhAoBDCX/JNLhJFEmNirI7zVE2NNppmbXnjdcsVMeGpnb5nLyK43/tAEgImAEi87foOjOwHaH0HqWwhSThoPF/7/UOOF5amc/3U2aK1oapbjBLxok1AKcGNapsXLzpAh1O+/0tvsvqzS7aH+yzrcwKew+IdHq3usT9+gmm6w4OTtwm+w4doRDlZH4CEneFNfur6f8KrD7/OT+7/CTKZcbD6gFQPSfM0/l9UTkAwbw75zbd+iR974k8zf/Uzl3/H0dWO9p05ma/jAvUCbz2jjTgMTKYp5x+sycsRWZKQpVksb7SiLCOQVSlNahLGeUKmNUWxpmtbRAjkJgHnYjNAKWgDZbFGG42RCvrLZ6okQYRoRvEeXzVY72iDxYlAKpKY0eUczkYKdtM0nK/XuKCYzJ7gmc/+MT44P2dxfsCHEwXjafzY2vojL/prm1NSrfn400/x3NM3mU1ylBLxaLyoo4LHe4tRGlJNGKSM8pRuc8xe1XCrrinKhrJuwAdc5/BE6cFytWZRVZwsVpyeLaNgLXhaayHREenc7/StddF95Szex5Av6+LFRSvd511FpafWcbcoGsv985YbmxkKECLQ9JNErRVKx56+kBrhe5R3UAihkRqCTnl3Jfk33z9mVcZoHHwgywyJSUmM5tg68ixha5rHRJAQGBYd7qhG+wFJGHK9fYH3336TZ/7Ix3j/7F2c73h++4/y7NZLGBKmyTYIxdKesWGuEHxgZ7rPywdfZ90tuD69w7MbHrP4gK3ZMUFZknSMo2VdHTLNInFhmm7zk0/8LONun1dfudjtAul2xUBKujagAqxWBdvbI7wL4KHTgWAkZVUzGg7BO7KemCCFxHYWESTjJCGTmq6qCP0ml2qDD4G6rC4jPbVSUWpelkhVoI1B9uCwmJhjCc7T2haLxwpPUJKuFzfa4Gl6ml7XtsyXK/7Pr3+Pz7DN524+z6c/9yV++9f/Gc7VXLD8P1zHf6TyZns64ZMfu8PHbl9H9c3PKOEMyL69ZKQCFK5p4x3aJEAgZAmD1DCsDc0gp64rrIsXP6UNIkDTdqyrimXRcLZYcraYU9UNto9/MSYhhFiSSBmBU3XXUdYN87LibFlQ1BXiQwMirePDkqWRvraqOjqfI8RFZI3oX3sfxNy3LSPTvBebSYGVmjdPBV955RF1C6NBxrrqqKqOREsGqb5kxMRjNfReWMU6e5/SHDJpnkIg2Vt+lm//3u/ymc9/HuhYdadsmGsYoh/AkPHE9l3eOn6Zl67uxgc5GO7sPs9X3vkXSK2Yjfd49/77+Pk9bty8BfoeAcNzw0A3ukbu13RiQDp+iuPf09RH/fEvYfTCCee2wpjYFdveHJEoCULSNBY1kGzeGNGdg7ce27V4a9FCMs1znI7YjSxNcU1HZzu6rgHnGE6mkecjBcFBnqSXNXaiNU1V05iSNM/jwvex3etDoAu9RCW4vjEQGwxtZ6OitooUhKoLvPrBEXzvuzz3uZ/k6o2nuXX7Rd5+/RtIYil0gf74yHr6Tz3zNHefvIqWRPah7xAokgv2uLO4po2sm3WNax3SGJTRSBNBQUZpXOYojaJzDtujm0GSZgnDYcbm1HJtZ0bVXqEqa0BRNTGRw2iD1gYhA97HtImu7ZgXBUeLFR8cnXL/0SnLsqJuO/LEsCqqHk4VKFvB2dqzO4qLW5l+cfcTz9hOFyBllM8LQWEV37vX8u03T/BCY4y6rCFDiMECF7u+lBJPLIcSlcTvh4KjydcZHz+JQLB3/pN842v/lPVfW7M/vcV756+zceUqfVYoAsn10V1O14fcW77B9dFdCDBUM25uPcvv3f9Vvnj753jHHnFwekizaLj97G3UMOVKYghhAfaUWm1xFm5z7//I8T0JI99zyBvvYN+LF8BVUTAd5jE6IcRsWp0meEu829ATzpqWwQV1QAhwDtfEpBnvLN5GjMh6veLkJInxSdaxLAqO5nMOT46YpSlX96/QdRlZmqJ8fO/j3CNqqqyPi957hybyNa11dDb+jq5zCKlZNw0/fPN15mdnTGZXePbjL/H+Oy9ju/LyIbvAf/xhYWv/wUV/98l9tAo426IAV9c0RYUUCmVUD8zUGKVIJwZsDA9wzmLLKrqknENoQ6o1idYXwkVafyHjjXEpzsb63U59LIF8QGoNErSJAV4xF9Vh25aum1HWDc9c3eGDozPeuH/AB4cnLMqGNokZphLBw7OCH9wPbDwziIgS8ZhKfGEUuUgjRCjOG8HX3lzy2r0FSus4Qe6NNBe6jq5zEV0tAqnpWevOYZREyNiVebD/y9w6/QtonzPorjJ57Qv87u98hS/+7J/ka+//38w3j5npHYSQLOYLfvmXf4mf+JOf53D9AZ133BjdQXjJzckzfPfhV3jl+OuQJMye/jyi26YoD5mkMUPWqwkhG5GlQ+QPAg/+5WMf7rU/s6YTD8izEctySd1YJkPRBxpEoZdWCi8Eha5AgPMhXmCdQ9HHgl70xONTTds0vPHue7z34IjNjU2GeYaUkoP5kh++/wHrouKlm1fJjWE63bjcibmQV3jXy1Asligga1sL9E4o52i6lrqnVEgpODh6xMOD+9x88mOMZztcu/kc77/zXQQOpeJ9bDDI/1Bq8R+isoS2LgjWxc6Clkw2Z/i6w3cOqo7GRbqXD6IfPhiUMuhMk2QB4cDaQO1a2jqmYSijSXNNkiZ9Ulw8SoVSeBvQOUilCIiozVHxTYKA7zo6H5DekQxSpnnCznTIU1e3eevBMb/z3R/GwONgY99WaQ7PaxbNgN1cXIgTHstT+zlDEIp7K/itV455cFQwGA7QWnKZO0D/oIiASSVKx68NM4OWkixVyItsVyk5Va9xkv0+V8rPIxA8u/jLfPWX/nt+4otf4sbmbV579E0+c+1PsTxb8fd+8e/QdRW/+a9rfv6v/zyvH/8+LxePuDq5wyiZsD25wZtn32Uz3+OKusWBmzOlY1xrunSbNtklyBRqzct/d0I7jyeIHgae/nM12e0vcOPm8zx8+CpZJtjc3GJx+oAfvvZtRsNAOsgQyuHoaGyLNgphO2xnSYXEAKrPob3YMJwLnJ0vOTw/582DY7Y2JozzAaPhgCd3dwnWsrc1Q0nNIM8vo4AuLKiX03IhCd5dOp9CiM475y1VU9HZFq0FmTYsmoqzs6NLyeBnPvdFjh99wGJ+QGe7aD7y4Q+IB/+jF73vGmxVo4QiyTOctaxPz5AuJllrrUl0Gg0cTmCbjnpV4WwE+CAFUmtkkpAkcfDhgscRcKEjdA4lFcJovPS4xqLiOJKuibV9CBE46p0nuEhOs7bFtU28OAnBQEnSccbsznUyo/jVr38XKaCoIzsRkXBWSLamCikfZxb17x0WyRsHLf/2tVOWhSPPM7QSCBGBqJLY284STaIUeWLQKg6wtFEkGhItcZ3HGIXtYLWe89rk77FTvoQiYeiusvGtP8ev/tK/5M//9b/I4erXeHf+fabdVX76p3+ag8MDbj/1NEcPz2A+4dbtK9w/f4sPbIFzJUantLZmXhwySa7SFZY1moUf9L3/ivd/JeP1f3rBgwlc+amCrRctQu+Qjze5dutZslQhRCD4jrsv/Ak6W6OlYX56iF1/jfP3H5Alhi7EeYbzEuU9WsZOmVZxyaRGc+tahKq2HtI0JTjPxmTI9o1ruM5ipGLvypXLhe77eQVS4OP2HU95H2i69g+YXdq2papKWtvStXGiLYVmvVri+t58NhrzmT/yJb737d/k8OD9y9f2kcob23bREe895WqNNIbxzja2qGnLKPAR3uOdQ4mojZHakGQZA2MIQuCDo+ss5XwRd9ZEk+Q5WTroIxGjnt4Lh4u31T8w6ncuSk+lijDPLigSnWDTDBc6rO2o6wgITXXCM/tbHN95gt/+/uvILKP1Ef7/4NyyMTVMhpEURoDGBo6XlrceLnn5vXOEVIxGGUrGeaYSiqQf0CEU3vcB0T7qbozWKCXQSvTMRk3XOuYnK3CC+9lv8cbkH/Hs8ucRCJ4o/gyv/4Nf5LVPv8YLn/oCv3fv13nhyoQXP/lpXuxbbv/wH/59fuLHv8DO4CqpG/HB/fd57taPUdVlxO5VjnGo6KoDDpXirFiSpJZwMOPV/3kH38Y/eP5Ew5P/xSO6oAht6GtwQ93G7pcUgjTbIRPxXjIY7mLEjF956++yrhpSJUmFpvMeYaOeymiNM1FlKrTi6v5eLFXbjuFwSF3XJEmCkIrxbMzu9jb5eBjTAYmtSA+gJQ7iZwjY4C+/f/HZNA1CSLrOMl8sGaQpflVydnqGc44kTSE4nrrzST7+3Iu8+cPv8O1vfoWjkwfYj1LeCMB3Fu8caZ6Bh+L8nGw4YrA9xteO0FrwF9mlgBc9p76Lx5UQJGnKcDxCKI31js522Lrt67x+yCMc1nfYro2a/M7iW49w8RIdp42etinxzhKEJ2gQSpAkGm962bN13L2xy5sPHvHGw0dRly4188LxWz9Y0jmH1AbnAotVTDO3zjPINeNBciHJR0oRkwK5oPn209leWiyF7qU9UemXJNFMfX68pmui0TxQ8b3x32anfIkt+wICxe17P8+3/ttvsveLns9c+0m+c/gVXHDcGD/Dw4cPOHjwgOvXbwLw1a9+hV//jV/jf/wf/id++6u/zTtvvclokPLXvvxjtAhWnWZdNax+fcLRP7hCcxzblDLx3Pkbx+irBdbFoLaus7hVgdYxY1YrjRShv6cEklSTbe4z3L3B2ck9NvqOjPQOGQQyRKZYnGNECpoUgmGeUJUFVRkJ1UVVsFwdo7TGeYsPewipohy8i5BZRZS2KPlYU2Stvfys65qmaWm7lrPzJZ7IypFCkuUDhIilktGxpSqN4YVPfZ4nbz/H17/6a/zglW/96Ive9RGaSZZSlxHoNNne6GkIEIxCaI3UGm8dtozhDMIFdJKQDwagJF3X0LQNztfxYZKC1MRALqFNBG82DuEVQkR5qQ4anRiM1jEXtHP9sZfSNDXWNXgXcXLFfEkHCCOxbctYKZ5/4ionixXrNpqaR3nKum45WlUsy3W8xPqA1pKNcRYzVYnsfW36EK/gqWsXW5whxj2miULgcd6hlYDgkVLQdY71sqWpLRdCeAcU4ZhvTv4Wf2z+iwz8LiYM0d//Kb7+Nz7gj//tXT5760t8+/C3mDcn3N5+kR//8S9w//59bty8yW9/5Tehh5Je3d/nV/75/871qaJ6fsKqhVOz5vTrVzn63+4Qin4CKz03/tKc3S9WlE3De/fmbG1uYftNKM3SvkwJFOsarRV5Zqhbzw9eeQM12GDRvEuqG7x2KEJsS3vQ/UXe9Sahsmo4OHrEcl2Q5xmrahlb08HTljbGYPrYKk4zQ9M2ICVGeLQMtD4Cp5xzl2rbC4lGVdeUVYH3jtEoR2vFINNcu3btUg7edQKvBLKL3MwkH/GlL/95nnr6+f/gole/8Au/8O/9pr338i8goKni6D9VhtP7D1gt1iA0Qhl0nmFGI/RwRDKbkW1tYkYjvJK0dU1VFJjUkA5yTJ7FHcBa6qKkq5t+GhrFskoKjNJoZcAFbNtFA4mz1GVJVZYUVUm5XlOuC1bLJXVbkw0z8ixjuYxiJOsDgyzjaLHk8GSJEJLJJAMRmBcVRdkCgWFmGOUJRkUxxwXd98L3GrzHJBpjRC+ZFkwGGUmi0Er0fW+FkoqqsNTrNkoCfF8CEGMQCn2PVf4G+9UX0GGAQOIPphz+jmK2Oebpj9/gYfM67yxe5s7Tz3HUxkgAACAASURBVHJj7xZt0zGbTsmzjNt37jAZT3BtywtP7dMsFzz4wYwH//hTLP/5XUIVYVxBeNI/8UN2/tO3GU5SvPO0rcXohNbGhSGlREjNYlFR1x15ZkAo3nrrPa7vTfjcZz/FgwcPOD64h1Yy2jwJaKUwSvUxSLErd3J2xtHZCU3XxhwuqbHO0nQtnYsnesAhRMCGi/akRyYGS8BdRmx2VFVFURQopaiamlWxpqgiXcMFyTsHJ1gv+dKf/lm29q7HhHKpoqaqZ+Xo/tTf3d1je3vyt36knZ4Qoum7aUnTjLPTEwZb24y2dxHoqC+X0HY2Coy8hxBFRlk6g40Zvu1oy4KqLAkhkOY5o9kUPx7hrCPYWJbYrol+R+/A94l0Wl2yUEKa0dUNWdsShiPaNu72nbeRhFYWDAcDVnWNt5ZUSj5150neO5pzslyxZcekSrE3G7MqY5RM7M7E1Owg48T1op4PIZAYHcFIwaMkpEPT3zM8ARFBsCFQrFuqdQsuIAO9/a43mItoLTzMf4evbv+XfPr8v2Ozex6BYP2W4Rt/wzB97gpX//RPs/En3+L16jv8wHyDG5t3efazd/ns5+NFWHjDn/rxv8z7v/YOr/yTFt7+FLQpF+I2b1rKT36T5gvfZl5cZ8dFzr/3PkpErOMyfaW0nJ4t2d6eorTm8OiYva0BTz2xT9sGrt19kVe/9/s0bRvTwNMsApu8pG0rWmupioLlYkHTNORZziDJubK9jw2Bk8UJZ+tTAo5VvSRZaU5WZ+T5iM57runrkZbmozmn61rWRQxUW6xW1H2oR103mCS5vODu7OywvbuPQGK0BqJbz0iDE1EGEtVLH2E41dRxIqaNYVVWbN9+kqAM67qmrVvKPtNUKY2UijTNSJKkXxAJwnlEAJPkZIMhruuwbXNpL0RKnAqYQY4OWa/aC4TOxvRxa+MFNkCQEtIEnRjqpqa1DXXjqKp1dGRlGZW1cHFzD4796YhPP/0Ev/fWu/EBIzAZZORGsay6HvEX21tSRGuC7GkKQsv4IIfYuTBa9adBTMkwWuKtp6kd1arFNTEPiX60j4hqTPopMCFwkHyV39r+eT61+Js8Uf4MioTgYP6yZv7yCP13Pkl+6zm6nUe8fOdtviffZJTMSM/2EG/tU7w9o118GuyH/6iBdvOQ4ou/Tnj2AxaLgo31jNVyzWCQsy4rrI8TZ6UUvhY0VSz50iyhs562qnjy2g3K1nL/4Qm1SLnxsRd57/tfp6kb7ND2YKV4In9wdERXl2A7hFLYAEfzNfPlA1rnWDVrdAKzSUxIPzw9pqwbkiTFOk/V1ty980ykTJcrvI8OOudiebkuy0hd67ooeixrAoIbt54mHUzwIfTGdIcxj6G5XZAIF7Nkf/RF3zQorfHWM93fx0rF4eEj5ucL1kVJWTd0baQbJNowmQzZ2thgY2OTjfEEOgfOszo5ZzAbYfIUYxKsEDjnwYfHhggf8DbKYqVJUSpB97ydIER/V4hHpkkSQpaSscG4q7Btw+LsnKoo4uWrv4Aarbi5t8V7xyesqwqlJAbPU1d3eOWdGMSglIm1uYjDL60iYUup3rTuQetelOwB4UhTRQiCpnLUy4aujMKryB7tNfjhsStffMixX6iH/O7mf8O97Dd5fvFfM3O3kBgEAlvC6tUUuAm/fRMJlP3nv/sRCFi1ZnH1d3lz7+9THt/j9s7H0fmER4cPGQ4HXL+6T1XXWOcYjce0ZUma5pRVzWQyQknF4eEhm1szghSsqoaXX36V5557nuKJZ3jwxiuszg7pup54IeKpeL5cUK4L8iShqhzvn5zyYL5mOp4glWRVVTRNzd5syJN7G0xyA3jKtqAso8Vzd3ebNE0oigKIhqB1WVI3lmJdUtdNH+OZ0q1avFDcef5TCBnLY2steZ6jtentqj7S9ULUF/3Iiz6CMxQYhRNweP8hZ2dznPVoqRgPRphZiu0cnW2pqor7dcPZ+YKdzS0GaUKeZsg8xVqg6WJbEvAuxC5MX0dLBCpJEMr0WnWL9Q55UQt2XdTVty2WQNOXN23bEFyHzhM2sh3Oz87ieL/vA+dZxmw0ous9rmVZs7ExYTpIqJ0lS4ZRH+8tWhp0oqISUQS8l5d2We8cINC9x7epPc26pl7X8fIVV2K/s8cj9t9V/F0OdmTDG9kv831+lVvu87zY/iV2ms+S+S3g3293i7p/S5Xe473RL/OG+Ject68hHngm4wF2f8jO5m2OjwvyfIh3gTRNKcuyn0tIAk2/gyqauuX05JT9vU1Wq5LDR6c0raW1lk4kXPvYZ3n1q/8KX5aRWCyitul8VVCtSwoTeO9oxfvzFbVznNYtmTa9ZENwMK94NC/ZGqbcub7DlZ0pk/GU9XrFfDFnOBjETk3dRB3WuqJp4zoSIqptpdYsq4bBZMKNp55B9iF5Td3grHvck/+Qkf7D9Iv/6EUffMDhSYYDVus1Sik2N7ZACIzUJMZE6pmSNNYxn89ZrwuapuXho0ckScJsPMIgGeUDMp/EGl1Fp4xzDqVU1OT7aAZWfZ3mvesTLQJV21I3LQFonGe+XnN6fkbd1CzXK6xrmQ5T9ramjGYT5qdz6s4iJDgEG7MppbWUVUkqNV1Ts7sx4o0Hx9w/XiAITMcZ01GEWtW264chF0aT6NJJE00AuiZQzGuKRYWzPcuRvpK8NJ5cLNOouxAhvp9CRO+5BFq55F3za6wmv8PI7TJp77LTfI5Je5ehu0Z0ZUGjj1lmr2LNisX4OxT5mxytDliuK6SXJNowHCVszAz7G57NoeZs8QHL1LC1c4X5fEHXdkynM5ztS4JeavDo6JBn7j5B01jeeONtknxAWVVIKdl76i4H77/O8TuvUNYVtk8esTaQD3I8mlXbotME07YoRN8ZUhgdZeBd13G8qrHvHnHzylWuXNng4NEB6/Wati9zOxsRjwhi/ti6iBd3BEeLNUfLNZs37zLb2OKCuZ8PBpG+rDTFek2QgizPIz35o+TIegRCK2wQKGGYjHIEsUcqLjoB/e5ltGQynZDnOeuijGlyTcPp+ZzxYEhrLaM8ZxgCWvUT1gBgsbZDK4MMEtd1uBB1F9Y5qqahbC2dDdRNzdlywflyQV3XtNayrgrKuuT+ieNsvebJ/V2SnpMfrOvvBfHC6qwjHRjS1DAhni510/V+3JbJKOuVoJBoAaj+ziIvDg66FlbnFdWiwts4BBFC9FxFcXml4OIxCCGqTy9q4sAlGPZCfGWDpVQHFNlDHma/iXeR1CVFdK8lBoTpxXxa0baOqm77ZBDBeJBw+6l99vZmBFszylOyTFA3B5wdeToHteijNNOMvb09rOuo6453332PT774cZIkYz6fc+vWkxBC3NCU5pmXfoqje+/GZEDvCd6TGkWeJmxMZ5Sl47WjEwQa3/W0Ie/xQZCnCUanSOd4enuTa1u75FnKIMvo2igbztKMrjcOtW1H07Y4AbXrEC7lbN2yKBo+fuNJ2rrDh5o8zwF4+6232drcZLoxAxGjTuPC/Qg1fRR8SbQypEmC0QbcYzuWc+4Su1B1MUncBzDaEJJw6Tu1zhN8YG7XNK0lT1MSrS4Xgu06rAgo6XECymqNA1rnaTtL5wLromJRrlmVa5q2Q0hJagyJGbMzm9C6jtVqwXsPDrnz5E2EkNiL1xpiOeacR10ERSiN0pKmaQkqcjq98wgZ23NCgHXRLyBFlL12LcyPS8pFFbVHF5tD8IhwYai5cHOHS1O60ioeu729MBpgBEL2D37o/f29LjyIKNNwIfpbvVAoL1CRzcdqVWKdRwlJnmj2tidsbwwY5op8kJBk8Y5wenbM0VGJ01OE2CBNE3Z3d8kSw/m8oG1aVsuK45Mlm1sqxnJaS1lXnMznEEBJw/YTz/D+9/8to0H0JSQ6Ni+MMHzixk2kU/zw7ISi36gCgbbrCC6QCMlzV6/wuaeeZJxHE/j25iZVU0RfLLAuC84Xq8tSxXuHTAxeKyrrEcmApz/2abRJ+xK1jAhIKSnLEh8Cw8mYTMe8qXX9/3UL+v+56AOQmIwkHZKYLA4EbIVQ0W/atJa66VgsFlR1RWNjIIMQEmNSjE57X2g0BkgBq25NVZXkaRLr317MhXBIGQNw687Reo/tvQBFWVM1NZ2zpFnGYDjEW4try4gK1/HE2ZnNWBdLjk9OGaQDQhuP8ov2oTHRTrgxzVCJYZAYyibW6hc5U+YiLdtHmWuWZFxcYpfHK8pFjetc347sW6tBoKUkUYoA2B5dGCsbEVGH/ZHrgT7xlAvujvf2Mjwhvo4IwQ0h9O061XeAPG0d2aAygFGS0SDlxo0ddnenaC3J0oy67lgvFqwWBb41uHSfNZEVOhikhOAxWtN2HRub27z3/geYxKB0Qtu1jMUwnlAy1tTX7r7A6f13WB49IEs06WSETg35MCXNNJ984iZSKL5d3mdR14g+HqdrO66Pxjy/d41RmvVS8VjqRgLGEiklTesQ/XvufUDrhGQwoHaCk0XBzv5NNvauk+Yx3+oi5ujqtauYJEFq1atfu0tN/Y+86A/OGlZtwdH8XQ5PFyzXBWUTIw2liEG2eI+3No7miY52oyTDPGV7NmM6GDLNU4apwRgVjRwdrNZrMpMQvCP4gDFptJX1/3nbczDbtqXsbWWhpzDIEGiaGogeWSVi31ZKSaYNTVtEI7aMdw8pYjdGGx0j20cpudH9KSBiYrvsNUBeEEIM79IqziKchfnxmuKipIkSzRgn6j1GwmQQd562ddBFlEU0qYgo3PqQwlD1l0LRtbgQaIMn7Xd8F/ylGwhiNyhAL/l1rIrqMhk8M5rNzRG7Vybs7W8yng7pWk9xct5nOlloSrqyoizXMVV9PmdjYyt2wwLcuHmN+fyMqmrIB0PmiyXT6YSmaUjTDBUCw9kOn/7iz/K7//qf0BZnVG1LlqcICcpopkPJS1efoFjXfOfgPtY5ssSwOxjypWefY386w/cXGSEFKkjSJKFJDEVZY13sxDRNjVQKnaR4NOerkmVZ84nnX0T0YOCLZHprbewsxqHKpZThIj3nR170/9X/+o/wzjEdpmyMBuRZRqoNpreRhQt6gAsUdX/zbmOt3bXRoKCkYpQm7G9MuX1tl/2tGZvjEamS1E0RAU9NwzC16CTBCUHjHC4Emq5jvlwzLwrWVUfZtjgCto0y1DzVDLKEPDHkSUJuEoyRQBIfQNkzM+mhsD5uv2XZ4BF0vSZ+kKlYWnkHXTQ6J4lGinjhXM9rinmJ62Lin+jj50UIaCEYJordSUYIcGobmr6YeUxKiL9X9C0eay0BifD9Tk+cel64/C8UoBfIwItTo24sTWuRMk5HR6OUm7d22bu6hdCKum8t3njqClpo1ucF9947ZGQdtV0yf/gqr6vAk89+AucDXdcxnY6jrMN6hqMBjw4PoiPOxLTHoihI04zRxi57Tz3Hu9/9Cm3XPVZcCgEarmxv8OeyT/NHbt3G4ZlmOYM0YZhmrMo1eiMlSxQOHxPH+4e7blsc0YBuO4vJc3Q+Zt1YTs5XJPmYp595Aake+5Jd77PWvT9D9A9C10VP7UVc1I+06D919wk+/+Kz7G9OCMHRNh2L0wVlWTMZjdnc2iabDHHes1quODo55dHRAU3bxFqUKChbVSWPzlf8X9+8z7qnI1zZmLA7nZCnCSIEMm1ItGFRFizqNn4WBatVjEhXUqCNRqnYMRIClPBIomiqbRydj6PtaGSGySDlytYUYzRlZ2N0p9YI1Xdh2ovAYXUpLhN9Wy423STFvGH+aEW4qLVCX5v37UkpBJnRhACNtdEV9AeO18ezCPnhxSyiSE+FHnlxMciCy5IvmrMCLsTAj6qKl1ctFZmW7O5OeeKpfa5c2ybLE5yz1FVDuapYnq+ZHy+Yn69wzkfrXWM5qhZUxZobz7yIkJLxaNy3CAOj0YAjoajrmkGa4W152Y3pnOPGcy/h25oHr3+X5bqm2wajwadwPD9jM9/gmfG1y1236RqOFme0xrM7nSB0/94KSRAd82Ws5ddFQUCgTAJJTtl6zhdriqLkqU98jsFk1p/KIQ49BbRtw9FRycbm1qUMQkCUjnyURf9X/8xP4rqGal1QNw2rdclquWZ//yq379whMQl101I1LXrWy3Cd5eHhI6qqRKtYduxNJ9zY2eLTd+GNB0d87bV3uX8+RxsTp3RdS9s02K7DOU+SRD373nTI9Z0po14klRgdjSoIDOFSpZcmhkGWMswHeAJV51hWJQ+PT3n9/YcczJfoxDBKNNNMk4foljKJoazcZakCRAdO333oiob5oyW27f6A4+oynCHES+yqjPGSQULX9bu6+JBJ5UN990tymehDCILAxS9Fpc5jmX88tglxamodnfNIBKlWbEwyruzP2N4ekyiJCJ5EKWSWUXRRiSqVYLYxwXtPWSSYsqYtzzl545tkowm7125DCEwmE7quZTgYkmcRxNp2HUkaAatpmlI3NTYbcOuFP8ry7Ji6PKdsGoZZikygywI/fPQuro5ZA5132ODRmWF7skmH7d9f3d9lBCfnEQZQNy2j8YygDevGUreO5bpAJYZPvPR5ssEkJqCHQNvGgGdjIi/15Pg43jFF7CYO8vyj1fTVekFTN6zXa7wPtF3DxvaMvb1tludnHJ8vuPfohNrFTo1wFuGjeaDwgXVdkaQG4yxt1zLIM56/dZUuwNdefYfgPc52dN3jT6ViqfGxG1eZDhISrRE9jtl2DakS7G1M2JxMSLXG9YmGXd2wto4sTZimGfvTXZ67foUff+Zpfv+d9/lX336ZB0XLiQo8JScMgFVZx7BdCyNlotJSAEKymjeUZ+tevvB42UaZQmz/XcgLnA9U7QUysM+fulzw8WuxxPeXx/HFlh64AN/K2L+X8lJDcnHRvRjOQZw2axXY2hlz9doms40hSW7i6+iifxgCWZpSiLr3G9Ss1zV1UcV/0zU8evXrDIdT2rZma2ubto2codlsHLOuhKVpW6yPphpB5Iq2nWXz+h3uvfZtThdrNsZjtBCYTJNtZJycLln1OqskSZjlGdrEB9j7mHYogHVZczpfUTYdg/GYJkiqdUUXAlVtqduO6c4++9efJM9SmqalqTuc66irsg9Xiw+C70OZsywjTR7zfn6kRd82j4223ncIEf2yb7z9NqeFoxtOOCfndLmgrWuGqWZjkKJChdIK0Qlc5zBKofvQsDxN+PitG3z7jfdpL5glIfRgVIcUhs3xmEnea9u9w7u46LdnY/a2pkyGKbnOkMGjdCQSrNc188WKdSkZZjltm5OnCd46bm7NuHtzn+++/YDGBqousD4vEFqhAyzXUfp8Yf0jCKp1E+G0Sj4OXBMRUCplXPjEdQtEhWa85Dze4S84O0HEkggB6nJaG1uQUkTqWmQIPf456S9KHXEZA69lbFHOxin7+xtsbE2YbW0yGBnauqG0NW3TUJUtZ6dLmrbDh2jEcTZ6AmInVdIuDnj7d3+F4dVn2fmJLzMZjQHBeDzuuyCBrmtIkhGr9RopJZPxmEpr0rsf4+j4kHuP7jEdDtnbmGBMlKEkiWFd5FEIphXT6YjpdMpoNIoGEwR113E0X1BUkStUtJ6mLOPgK0DdtJgk5YlnnseiKMuSpmmjkE8qrI3pOCF0/dfkZaiEVhotPoJz6g8eE/FIXxUVrdB88o9/iYNlwSv/z+9AEHSuo1ysEHKLJ67fZv7uGySiAxnjEfM8RSmFVpphlqCUxnYNvmtxXYfvOrx3iEzGfrASJFrT1VGgliSaxsG784bupCT4QK4FAxUYJ9F47un5K9ZRNw1JfwQu6wojovOn61qSRPLkzR3u3t6kaS1lHSe+1nq6NjJ1jNG4Ls4X3AXoP4RoLkkUSklkEAQbV3Xs8rj4M73RJNb/MqbeIRHS9L3oGBahECgR259ByFj9C/oWq49E5A+hSxIlyBPF1saQjZ0pu9d2mG6OSVKNNpqAZHFecHK8jMb1EHNd66qhaVqcdSgp4mwiBML6hO7db3F04xpXbj2PEKqXH8f27XIxp1YVSZKSZTldF0lkAclgY5eDgwe8dv8QpQTb4yFpZsiyhBhdr0iTqLm5oOwbY2is5Xix4s337tEKhbPQtU2kLuDj1Fcbnnz+JfaffIHDR48YDofx0toXgcfHJwxHQwLi0iIYvx9ti3/Yxx8av4P3WKVIjCY1Glk1jPIJbt1weP+A4OKt+eGDB8xGOW+enfLk9RtMNrc4Wy0YKcP2ZMJ4PLw0jj86XtI0LaMQ+LM/9hI3tzY5WRf86+98j4erAilh98oWxXyFk7IfJKUUOuN0bZHaUDUFe5szxtsz7n/wHn69ZDoYYJSOo2haWtuCECzqmsZFCJRwLVeubPPSZ59nPIyZqN47lPAYKYlsyQ6Ej4jyENuG3kYIVedjvdt2lrZ1rNcNdWPpnKeqOlarhqKKu23bWeo6Ig+dDzgp4uzB96FwMpAiwXuqAKrv5eh+9iB7piMiWhKHiSbLNIPtKbOrG4y2hggtaGxLWbVUVcPJyTnFco1tHU3VUlY1bdOBjFoW11k6a0kS02ueHKtHr3P11h2UmWCtwyQmnkpa9SwfHduF/RyirhsePnwEKuW8Knj53YfcvrLBOFWMRwPSREZsIzFeqSpLhJA0XcO8qHnl7fe4d3SK7I1EUms6G5NslDZce+o5rt/5RK/YDdR1RZrmGGOoqhJrGxbzjtF4ijTyUs5ijCGEQFl+hOHUJIm7pwgpbWcRUjHtk+iaBw9Q5yuq+TmL1RJXl3Q6sD0akrRLaGum+YDQ1kyHGduzqO+uOsc7Dw4ITct//nM/y8996fMI51it1jxz6yb/yz//V3Rtw81r26ymA9565V188Iy3d/jB+w/QgwmpEAzyAffu32d3Y8z21Ws8fG1O3baoTD0GrPbdlUggjkjBRIJRgu++8h7BR55m/ANr0iTKjocD3UdADtmYJeQDQ5Ia2tJTVD3MVAtMoi71+LIXqVnbRfVlbSmKkuWipGss3sUWbOf6EIGui9TluotIkTa2eZ23OB+zXS9iMBOtSIwkSw3DUcp0b8qybXnz7Qfcf3TCIEsYDAZIJXn6xacjPazponisbrFNR1nVNHXLellQFRXCRd8sQrC5JXn7B19l94nPgTQYkyCVJMvzqJFpmzgQIk6u9/av8OKnP823vvE1FuUaoQI//OCAjUwxzFM2J8OYHeahriuSNEPMl5Sd5d3DE+ZFS9D6sp3rnUUgUDrl+t1PcOWpj1PVDV1nUUnEzHRVHfX2yzlSGyazTYyJD+5FLjHQQ8I+QvdmOhnHC1kQFKHG+p5C5VqwLbenQ2affJGj1ZymbVABNoc5EylYa43YnJGnEfVRVhXGGI4WBR8cnvLirRv8zE+8xDhPwQVEZ3lh7wpffv7j/ItvfJP3X7/POE8ZGANZhnEOZTuWxwesizVJlmGkRJUNW4OMwbUbnC/O0UqRZwlKiJ5/EhgkCYPM4GzLdKD5+b/6M+ztbdJZj/PQdTHYq24a6ibi5Mq6xnY18/mc46OWdR0oy462sbQ2PlRBEI9jJZFGYVQas7NkpCFLbQhBYVSCkxDSyATSCZi+7FRK9DEykCnBKI/EBq0VaabIBgmjYUaaGbRWaKVIMolOokc1iCibGA4GqESRJCaCbLs2pjaGQJZnUSUqNU3T9caghvWioCxr6rrj7LRgVbyNTm7inGM4GqH7gV/TRPLEeDJjMh6jFAzGIwbTadRmVSuK2lGtatSy4GS1ih0WFUOUnS9YFBWN9TghSfNhf7G/kFwLstGMnSeeZTDdpm7KmFTeekId08ad9axWK5IsZzKaRNArxMhNpfsSMn58JMLZMM0idzJL0RJcZijWRfxDB09XrtmSir2tIS4M6KyjbTuKxRnFakXnPHY6xSnHII3ywncPTzg9PeE/+yt/kdkg73HYoBGMM81Lt67xG9/R3D865cmdLa5szKJzSXi+/MIzlDYwr+rYHRBwa2MKbY3QmrVUONuR6JxMS6rKghCM85RkVTFIDX/hL/w0jwrJ/F6B8zEJ3WiN0YYkzxkOBVOt0VKRpCZq8FXog+HEZdqidbHGj1zOjqqsaaqWomhYr0vWq4KqqCnWVUQV2nhZruuGpnVYG39WICMGRQp0okl7hqQxcbMQ0iJEhTbRO5wZwSDXJInCpIbBQJKnhiyzpJkiMYo00ZjExIWXaqQ2pFmOIJBnyWWqyvb2lLKoqauKK1c6zk5rHjw6pagmvYQ6ki68dQitSYzCmMh/n82mTGaRY392EnA6YXtzG+lblicPOZvPkaKjbZo4dXYeZUwfq2T7iCWFUJrxzjV2bzwNUrNeL6hPjtjY3ERKRdNZQk9F0DohTUcI4vs1yOPubrsObUwExpqLe9OPuOgTE5mTZDmt1lg8SsB8vqJp4oTSC4mr64jYdh1FVXM+X9I5Tz6c0blYxyYmZd053rh3yJ39PT7zzFME6wgmOuMlYJuaK5Mhz17b5wcHD9kejzmdLxC2Y5Sm7AJboxHXdyIxq65qusV51LvYjjQxKBnTzJUQ2LbF9ta9sqqZ5Zqhsgjf0nQO74nqPNvz4mVEU/gQhxxKaPDx+L1woUWgUPw3U23i8CaRceglQGcjtkczrt0wJFpHslsSI20ugnN8sDhrcV1HVVS0bYd1Ha1tqdqWqo4PTlPXlGVJUbXUa8e6bWnbeCG1NrYzs0yTJSlZEl1lIQIho3y4Z82kqYlCtESQ6EgmyAcJxiiUgmGeoAAjNNe2FW98MOf0pMK5OMmOqtgo947ZATE34MUXP8WD+/dRKM7OTvqSI0HmY3zykOLshHwyIuCgqmg7i+ssCURCQjZm+/qTZJNNzpcrpIjBfNZa1usVSZLSOUdrA1kW0yRdiLLm6XSGD/EeZtIYsnyxw19can+kRV93luloGFEeNZgLBF4fszdv5qyLNa31tM4SgsChGEy3MGkaE+bShI3RgMk445VXq0ntjAAAIABJREFU3+LRo0P+5l/5OcYmHtXSR2SfEZIECXXDx/b3+NoPX6cJgslkRlUWrKqC9uAhs2HOaJBhtMFdBJ4RgxRSJchS0+P1+tZikFihODk75+b2FLtasDp9FFFyQvaTUYNQj7sDUmmk1Hip+qGVR9BLL3rtjAA67+mcAysIRFmrDyUXuU5CxDRFLaMgLRDzqi6mxpIoVw6Inp+jyUcjppsRJmW0jF2sRPXBaH2L01rqqqTtHCrNYkh18/hhqYqKsmrjnWK5Yr1Yc3JQ9WnuvfLzgtR82SaVKC1pLQiVItSU2dY1jEmoqhIIrJZLlNIkSYKUks3NLWazGU8//TRnZ2d0XcvBwUMOjo8RyiAG/y9pbxara5bed/3WWu/8DXs8Qw3nVI+O2x3b6djtxFFMIAYcnCARSFAYr5iEEFcJdwiuuUAgbuAaBRQpEhFESXAY4iREittuu62m6MndXV3TOWeP3/COa+LiWe+7TzsoKFXV6j5dVfvs8+3ve961nuf//Ic14zhQFDkq82Q6A2UwRcHZ0+ecPX4TnRWM48A49HgfknBdMYwDwzjgvUJnBVZbYgNPn77B+cUlZVkK0zITRFAn46hpmjgcDp+86P/7v/X3+Hf/pX+OvJCp2HlPkVdsNposK6lXW4bJcegnRudBaWG8+cA4SubUdrOiLjK+88ELfv03f5c/9Pw5P/vOM6K1KO9ROUL6yjI0keg8l3XF8/Nz/u/vfZ8/9Yu/wKqqieFUPHjcxKEfyRgE9ssMeV5QJ2lZZjSFnn3QDT4GbrsRoufRdg1Gix1JnNP3NFqJu9bx0FKVhbwWY4TMpMT0yZhsKRiUkg/ytXwjrQ1ZluFDSgbUCqUNLvXhuc4TDi+3RQxidGRTMITIahW71i7rdElbTBrbROERyZ7w9rPcoHWHVpEsE9hyvbng9MyQ5TllmZEVhqIUmoUsvpLoJQS8nyRToB/o2p7uONAeBg73LV1rCfnEcZLN9TROHI9HTFawSVJDYwxKaYqy5Oz8XG4h77h69ZJxHPjSz/988ikNWOfI8hzSMGxdwBjB4Nv2uCQ2JmRYDJuUoigaiqLk9PSct569w+npGVophr4nz8TG3Xkn2bN2So4ZPfAnPlnRf+2b32F3HPjVX/zDvPPkXK5MP6FNRl5CrQxVoymaQNsL0UwZw7EbyKoMMsOLfcs3f++H/M43v8U2Bv7Cn/wlMuUharybyHwpp6wW1uY4TmRR8Yc++5y/8+53+Htf/wZ/5Gd+ipPVhlyLD05uxJIjeIfWgl/LIsejeeBzm6ygD57f+cF7fOn5m1R5Rr1Z0axKEXwYkzaqQWzozEasvhEe/czVC0FCAsTZzOMBN5lFVCHYttCngxBNifOSycz24FIgKtEpZPOcJRquQqUEDpKdhfwVQWlC4uJIHkAKaA6KyaV8gAgh9KLhDeK25nxSfcmTDTOH38iyMU+ySG00eZFRlTll2XC5Lnj6lqZQikDGb7/7it3RMk6k/cdA7Wp0YRJUqBOjM1AUBduTE95+9oybq1ecn12SFSW7/YGubanrOmkw5OFo2yPOya9d11HVFev1Ng2/nryoqOsVT5++yXpzQgyBw+2NfGaJsGa9w059KnqRlNq0VP1ERf9f/cV/n1/7h1/nf/i1X2cYJi7OTnh8ccZ6vWLdNOJvaDTtNC1LnvtDx65rORyPHPd7vLW8vd3wZ/7gF/nM40ueXZ5hkGtfzdI6DSqXjCkpjsDjdckvfeE5Lw4dv/6132G1XvP08oKn5+e8cX7GuqnJtBJs3VuUd0tLMVnH3Rj43qt73n3vA56uS770xiWvDrfU65qqKDBZDkhcTt8PmKpMkj7p7wN6Wc6pFNM4tzrxNa67TqqpgKKYF1Q+LFtc58XXJ/gHu8IwCjYfYmJXJkqDQuNd8gDKdHKMSJ75yqCy2WY8S4ouk2SN0sfHMOtW5LZRRJQyaXYRPlGwsvxy1hMRy0OlZTifk118EFq1ykrIcuomwzlJg2yPR/IsZ3tySmaEaRtCpK4bQvA0zZqLy0cMfcfLVy9Zb7bc3N5yfX2dojXXHLuOu7s7tDKs12u6FJ7XrNYcjy0oOD875+T0gvV6i/eeu9trYpA5yIeI8y4ZQ3kCYdEfCIT5j6ci/GOL/vn5Of/Or/4y/9Yv/xIffvSSD15e8+HLV7z/8iWvRkvXSzQLRmGKTKIUi4KfOF1x+fYlZ0XGOjOsmxXDMHJxcU5T5skhLS4npcrFANbUhZgrZRnbsmZnjvzUo0v+yOc/x711fHS/50fvv8+3vvt7eC3ksEIZ6X2Vwrop+Zo7SqO52Kz4lZ98xqouGYNLwVwGF+bikv66zHTiuGspziT0nsllOhHbHmjC85ZUPnSXGJjWB2Iu6iytJIBuhtKEpyO8+kzLQxSJDOlDnFVVoRDvzsQ4SxHwETt5MDqR0B5em3cimsh0yldN9OmYECGlRG1lUntjtEEZlW6djIigV0ppgtbJXUThtSHaCRdytG5km22MZIvt7qmbmrreEHxgtVqR5znOiW3LqlljreOjjz/GGJMsHiNuGjke9gQih/0O7yPX19ecnp7x5ltv0Q8jWmnW6zVZljONPdddK6f6nKv1mhhcZhGDygzaiPGT0Vna7n7CohcHgIByjqcnGx43NV883XL75J6sKJYId5TYLkct/9/7gLIO2/XkKatomxh8RPnAFluJMqPwhTAPU5tjlEZHTVM09ONAM/Y8bVa8886b5AmXBUm0wAdssEI11lqkZonP1U8jU5BUCxc1dVnz0csd+w/umOyEZERp8tygYqDMs8UfUWmEwlvlVHWOBspChDBKa3Kj02JLkSXFT6l0YlfGJHR4zR05xmR5Iu2F1kIoM+Mwv4lye0QJfpMCBkrpddVaBPMzdTmEuFCsYwgEL7oGH3n4/TFCkKgb5+R9Dyn9MMSY4nLmFHeNQi/DvUqfqckaNmefw+Q5hffEODL0I7c3NwurUcLwWCz5xlFoD3me07atoCx5LvaP3ouxU4g0zWrRIA/DgNHyUB+XQVQAAGVSIiJaUJ88I8sK8rxYvu+cOSut26c46b2dFmGzNkaKLEaaspLiTmomH2NaVScH2uBRXrgWRQpkMHn+YOwaAlkmW09vLb7rRGOKcOaLsiAO/SKMVjGAtejcYJwjS9eYtMAaFdNyQsEUIS9LcTUYB5SPFNoQUDyqa3ZBkgV92g/YyXPsJ9G5IoSs3JiUpSVdfZ5nyete4D6t1YKnG60xWlHVBXUpaJU2qT1XyfYvS0iS1vJASTciWLMuUIjEMARSy5SLQS1I4ENqs6pChmmf5ojF6cIL09N6ibaMmDQYkk7IhEuJAY+0NkoRg0t6XtEx+/kGjvJneid5h8FNFLlICIWOYOnalqsY2G5PKMuKcZo4Hg/CxEx+STHC7d0tPgSaZsU773xWTHit4+SkXoRIs7BeYp0MZVmQZSUoLfuGLENpiYcweb6c6gqBiotK/G8EdiYNzJ+06Ceb9KI+2Vdo1tsNeVHiRsFTQ8omDVE2m1qJ97jJ5dqchzZx+NDCtEtDXpHIQcFK8oR1VjaOuRRXmRt0EpjURUGhNZVRFMl7Zo5QjJYkPlUL/8KHB6eCLDNgB0y6/u1kpbi0ToollbwRBF2Rl5qhVDIs9SmkF4WPinF0uKjEIgSxPpSuQNqfiCBHSgkzUuSKKWFdRTIjqFNZFxgVqYpctAFaRPWZiZjcpAdDHgI1l24MYmiLXmjK3snrrECAGUVyALZoNReAFLSdHEWRpU1ovtCXlQlkIaIwy3AsRWUIYYRQopSiLEpCgLY7cjy2OCuGt5N1HNuW3f0t3lqIUJYN6/WGw+FAXTU8unjE7rBns92w3WyJUXx5tNIMo5OEG2MwmVAhIoqsKBIkmTTMRSk3rBZ5pk8HoDESlJ2ZLOWHfdKid46oNdZaUfoHhckKqjojGHkzfXy40qqsTFdcQGeC9xnxy8NZR11VsjbOMlQukekAvp+ww0TXid9KVUpIm9EN7aGjLnM2q5oqLzCZSdevUGSjSoG5aSBEKyYvbrgRWVSYPCMPwtF2uREkJsZlkAw+MEs9HqxN5OYI6efTyhCje1DsL8PqjPEo5rRurXhwXvAAAT8FslxuTenfLf62kxRtLTbk1tmkLJJWDSJ5llycTZZENJrVqkjIiSIvCunPmdEiTa4DucnxuVkiTzVisLV3rTjEpWIXcQbkSuYNpRTRy2yjk0sERoKjZ9mjSeHWwXls0rVWdUVd11ycn3Lcix9llmU8efQI5wObky1N0xAVXFyek5l8mUsgMgwOn9Q0Wokpq9bzQyBERR8jeVGiokDGASiTBhoSKpXLA/OJiz4kC+VFARQfBri8qiWlOsiHm2VZal/ktJ+jEwGm0eKDo2pEw6q1QufiVIWXSdwFGdiiVssprVVCKFDiUV9LQEKWiXHqrI0ki2D0woeR3CsZeOSDk4J0znK739OOE1VdS56S0kxhbgvk5wuJez73tsF7ggpLHw7pn4cfzy2dve2BJet2/lqh5M72IvLUSvxnJMqRjlJGECitCUosSfrjmE4yC1GCKXT6LCLSWnr/gCJhRF+QGy03SBLEF7lAlJnRGBMXf3mVaexk0VqlNMcAGjxOYni0JiphWCrEXlGnG9sypfikiaIqWK3ErUAphc4MzWpFWVZyCATPfr/HhcC2WadjAoFVlSYrIsNoBWo2Eq6XZ2LXV1UVRVXKTOIjMDMqQ+rjBcUSsU4gRPfJi15QFin6OJ9eJtnegazpeRBNWCvhZjFGcUpIhTsNEzoTn5dxEgdjbQxaST6VMAIto7O0Q8+h7zm0R2mpiMnwKEcFGTyzedjKDM7KhyBOwuJVY51LiIVJKo+IC57bYeQHH19xGJJ4WIsCarKW9WqdFj+aaZKHIi+y9LA4MTwtS/mggqiZdCrgWRM6Cxrmov/9WtkZspyH2yUxI3oWCeL81en20bO4BJV+FlngyGEgvbtS4IJsplUAPwU677iajosDg3xPL7NGMq9SqAWiLcsstQjCiSmMoa5KlBH9wNnlI/nMjMEgf/6EzCX90Es0pnOUZYnJDGfn52w2J0QtrnHBSgrl7GQwyyYFNMgxWca6EC6NMeJWbUy2eBflaQs8DSOZyclyQ4hz05eGfSODrnWfIokkBHnj5ddAlkk/7tOi4/cTewRXloL3zkmChwo468i04bg/0PV7Pn55RUDz9PwxRV7SjSPWi8vVx1fXvLq9Y3fcCzsyqxcIcH64Yhp4vJdsqxCDxDPGuDALMy2uyAERdfSj5aPbHTrPabI5fY/FR2XsB5z1C/57f3cPJHG7dygiWS720FoLL6lerRKiU9B3PevNSgbXPGeyk7RWye8mekFOVCriOZIzRImkUa8hOgJZxtTC+bSljMlZWW6kh9zUh9tkvlnmX818q8wP0iJTlE21UgofgDQsa+1RKhBjcozOhL9TlT0n559DZ9niOZNlOVUFwzCQZznjMKTIHKmLk5MTtNaUVZ0imUoOhx3b7RZA0mZS8YfgUFEcHuCBOyOJKYaY3huJDCopslwOuDBTyFNX4q3k+n6aop/lZRKFKFCcLADMcoqZBEOFJIcjBaXNlnchITshBPq24+rjl/yNX/8a3311y7/6p36ZL7z9BodOhOf9OHG93/HtH/0IguftR485bcS9eI5AD0EkgjO70UXp3eXvXeJvJGxdK4ga5x1T8HglzDybsO75jQ1BHJRjkNS8JXo9il7VJbIV89ciWaf3t3eyiEpF+OqVfM88lw8lz/NkomSYQx5OT04SMU5W8EKUKoWCEYWlWlX1skm13qIJskyLMcn9HoKCZdsq7dIsnn64UfTSnj78c3mQ5odhpicLopoerFS4JvFgQnCMY0+TF2kDK7K9PFcPXjQxYq2Idqq6oSgK2vaAc5aiashMRt/1PH7yZPlzp2mSbW56+KZpWnxtjBHqh1JmKXitJLhPop4eZpJAIKsy0RGnmfITFz3wYz29dUKqUt4vA5JK17gxQhMN2hMnS66MuA4XBUVVMUwjLowURcUXnr3JwUXe/+hjmipjf9hLJLqPvLi9lwVFU7MuatZVQ5ZnCVXx6KyQFX980KNKb83ywcoHkiVBtfBqRisnuNKabN5gvtZq6LTJfLDahhjl711IN4APqBjS5jctqYKsw+e/Xj9tZ17IZCdC8v982Q9LYc3vW56JxaDJDMM4sFqtUSpxXuxEpgzr1QO/fRiHhQ6tUZIEmNqUZeubuDuv3wDz63q95VHLQZVwp4cvXVq5EB2Hw46q2SwBCUUh/fY4ip1jZgyZqRNuLnyozXpLPw6MfcukDFmWJ+QJTJ5TlhWgmJzHebu8Z6+3i3LIKLwtWa3lfQmpnQtJMeWcoGd5ni9w6Scu+vja1fnjg1tI20sIIVv+vVZiwCqekIqirCjWK7RRFFPF2I9kpuSP/8Ijfu7nPFZ5pmFgd9hjg2cMgfVmRVXk5GiKrEgSQ8m00jGZx84DpWLRsLrwEJH+etsVQmAKnv3QC7z12nU/D6/SMjDv8Jc3XylpEXIUIaTWIypiUGijhDL7WqHNf95D8QX5zvF18Xfyv5HuRnrsufiAuqzxbiIGGPt+Ibkd9oflNVtnBckxRXJWUJLaUUpy98wpGoaRum4WUySioGjzPCPtk8Il28FMGYgqMVT1w0Mc4bC/48nTz5AZUJVanMSyLEsLwYdZpikbvPdM1lGUFXmWsd8d2GyE2zRzZF6vraIolnYnS21USB42c21ZN+GsTSFu8lDnSGfhnKMqS2JRzJ3rJyt6pWcW4UPBzyeET6aoMQrkpXVaEGktG7OioNisF9/LyEhwHh8KumHkcDzgiaKiX20JEUoFvbPcDAPaZET1YIMX09WmzUNrFVSSHceQJnm1LK1iVAQXJNli6rndHQhVs2Dpc5Gp5T9A4iDCw+wQgvuxYpavilivUo/Na39umju0Wg6H113K5uFx3jAu6E4MwlIjudMr4QTNydczhWGmNMw0DkhtnZ2IMdB3PV3bCf/HyVZ2p/eLEZI8HOL6u2pWGCPLnbZtWa1WqKoiRhnap2miKIrFnKlvj3j/kPU6836MkaLs+46maVL+a89mK9nB++NRlpk8uFxXVbUUviAyPgED2XLCv/6eynsUwXn5zCdPTMieQLTqx763c58GvUmn02xlMaMVwftUJjP/JCb+h0CXUStMUWLKQpZVXq7B2909L66vuD7cs+tbQog8Pb+gqiueNjX7YeDm44+463r68Q576lmvSqpVSV4myRti8+Bj8tCUqkvFkV51FAqAtSPDNLBvO4bJkZ82mLqSQtWJLpF6XxVlL6Gj9PWvX7UPpk0sv5rlOHm4HebWIfiQBuiHQyI9SjL2oMl1lrwdZUs6Z13ND9HrrUqMqdA1C6vz9VvJ5WKhKAdAJLqQOCpmGXZDEIfpqiyJEY6tJPeBPNDt8bjcoAKxevI8l+VRGkifvvVZTs4eLT/zPMsURcE0jbSdOKJVTc00TZRFxXazYRw6YoqyPxwOgoIptZzsWovHvocF+p5P+bnljCotPfU8uzwY284PyLL11p9iOTVOA0qZxWbBB3kjYyqamJCELMseWiGjyfICXRZpje45Hlv27ZG79kDnJrpxZJgs7TAwWMubj5+itOJ6d8/u2EnDYQxORVo3sQ2e0pToOLv6Jkpt+v7O2eSb89Czei9b3tE5bvdHnI/oELH9mCgLD8Wo9cyiVCmx7uHk+P9uWx6KcenrX8PkxfZPtMXzlpd0K2iU2J0rSUMnzKFsSWCuHk43Ea0sT3Na2yfPy/kh0gprZd4IcbY4tA97gkRFCEGMbn0MFGWZnpt/9GeaH6T515k6MtmJFx+/x8Wjp+kGEu8euYRkeG3boxRVmjf6vsf3kSLL2N3vEOKcpe87oXWkGWWGOYOXQp9bH5h3B3G58dNFLUzQhOPPwEFMcPmMmH2iovcIquG8SxwT0U0Key/1wekhwCjZ3ilQeZYWR3Jt2Wmi7VsO/ZGruxs6O8r3zAyDs/zo1cfUVc398ZCGzonBDoxuZEKSQaoYkqOZbFvnbalL4u/XN6MCZVrhWsdAO41iNEsgxARnvWbcH50Yxs6H9wNfRS/95Lzt1EYURjptQVFCbZhbLhHbJCiOsBSmfL95WBQv+nmLupxUih+7VUj1HhB35Rl+Fa/LuTVKJ2H6egm8e5h55m8SQqQOnuDCw6GVQI5ZyfX6ELxg+wsSBP1xj52mHzswsixLwWdi2TcOA7v7PXmeU1Ul3nkObcs0WWIUtG/VrGQh99ohlec5KlNLcc+vIU+cLXlI8uQzqlIWQtqLJGRnPmT///76xxZ9O3RpuJEfrirrhQIwa0VjSuRWKQo9hIBJxTUNA2175NC12OgZgyNmirEXddAw9EyygmRItFTnvYQpT45uGNkdDhRpOWWKSq45YuLqSHG5GHBR3ph5sJ2SPrYdRu7bjmx7Im0Mc82ns0E9FCTp9Uc1D7fplI8JRXAQVUwkuLmE5eEAaVVUQkJMkt+pdO3OqeOK1B5q8bZZ2qfXXoPU6TJ0YIgEJZJFlYbhmW0m85R64AfF1w6A+TaaAyKimFUtV0WiHmc8JPTN35Moqec+cas0Cu8t7XFPvdosX2cyk/r+h9/b1DUhBo67nfCSjGG1WrFebyjKQpyHEZSNIP6cIUTyrEgbXyWxmtaS55aqrB9midn2oywY+p6Ff5Xes6Ioft9S8J+w6A/HI0UmZKjM55L+UZQYlS1XuhB+ZBibuRuTG3GDEylY19ONAzf7e25391hnWVUSuVIGz2G/B6UWJXuWjZhMWIwuRo7DSDX0FMc9YSVhAiHICT8lmMsHaWVmiMt7jw3iYvxqd49XmmZ7Qixy2dzOr3nu2Q3kyYksJB7P/AFmSkQVKsSlDcEnKxTmHjxNqTEyXyR+wYrTayQuRW+UwmAIXqWoy0RKe62tmgdd5x3oHJMABXlh8j8zb+X33w5xWWwplEpidFQiwcnXzZQNNS8E5h8hPGD1MaFdsj2X3zf0R+rVZnmf59N3NoVarVZ4L0Kaol4RYovzjvv7W4Zp5MnjJ9RNk7blAnNrY8QGRulFSGR0xrE9YqeJVbNeAIz5JsrznCylCs63xdyOfaqiD8FhPVgfKQhoSyrwtDZXRghUaUkyF11nO9ErjhPXd3cch5ZXt7dM3hOCl6dea/Iu59CPBCXDaVNWhOAZ3YT1FpSntT15K4iMDZ7T1RqjJcvV/z5kRfp7ScfrrWUMkZvdAVXKniDYaTmVl+pRSpAh5ijNLAngtdxe6eQM1uNS5I73XoapINRbZaSgluZ9bmEW6sDyj5fC8rMaQUVsDCLg8CyQ5lz8ISW4yICWkLDUq8MMr6ZhPBWmi+Kvo0nODkSi90s78dCayuuNPNw2+nUCURTa9OzorIC7u5ecXsiCSSU0aGbLZpn00sYYDscD1lpWJ+fkueF+fyA4z4uPPpKuoa7JipwQIqdn5zTNOumd80Q/IQU1iHWKhGhrnLVLRzH/9fqmfn4YP0XRy0paK0XMssRtESPXTBfo6PFyUS391DRZdvs9o5twPnCcWm4PO6IGHy2n52ds6oZ1uaLvR1bFikPf0tQ1SsOmacizAu8Dh64lKjCjXpCjMs9kNf0aLj8/3c57nIIJGL3n9tDRO4euNWaBG8MyEM4f5Fykzkv+1TyUz6ZNc3GFxD3SSqUtqxCylBHimrR8cXltPhUasPSfYUFw5i5D/zhqQ3K6jImRGZMFSdTiSrY0Xeq1E16eqHn5po1Op7xZvofKCrQ8BtIOTja1P0I1IcHNUuBquQGWm4AIWjF0e7rjnqrZAkI777tuOXX71HJ89OEHfPDB+7z99jN+8ie/zKNHj7m4vGDoe1GwebfsMe5urjnsdmR5QV03VHWTbNyL5Xvu7u559OTJAqS8jjK9PgO8DnN+oqI3id+ikoeh9x6ntXgx4uUEChGxexOWWzt09HZg1x4ZneN+f6DvBvI842S75enFI86rLcqBygyXxRmDm1BacXfY04aBs2ZNuIzo+xv6YWDyjrvuACpQVwV1UsEzv6ZU9JNzBC2nYzuNvLi7TVi/OBGTQJuZ0Qk89LjMG14JZtOAccLiY94VZIbBiv2edbJhnTtorTU6U5RFQVkWKCJ5kS0PjUl8nzDTFkJidIa49P5z4TM/nvPVnW6cwLxcU2m2mgv/NQTGe3wQcqAKc5TPA1Kjo1oia4SnnoZoky2n/TD0ycfnYZj16bVGDff3r7gsxaWsyHOOxyNZQlAmZ2kq8aX54MP3efHqYzbbDTFE3nn+DnVd0Q8DYy8SSJH4QYweZwc6bxmHju12S92s0GmJFvDc3Fxzef6IZrXCOp8gbLdYvEuLqBc49xMXffqpBWuPdvmQVZ76Uy2sy6gUNniOXcvN7o7BO8bRMYwjWZ5xcnLC2cmWdV5hjwNhDIyj5XgQ3k07jvR2YvKOCYvWkZNmhUlr9xA8hxYuT07wZIknI0NvSH36GDzeS+7svusZJ4sxGkaJWSfLycoSo6KczlmGRJTr1LI4grOYGNARIS+lPr/revop0E0Oh0FlFajE+c5msphiHAKq7yB6iJK7WuR5shSfiW4C1ymTJVRMTme12IkotBIatUbh44yuBEJSXqW2Pp1wy0iRvn1ChnhtIFd6uSMMgJZsLRdBeQWJmQox9eeG12eFeS+jtaZvd9hpIEZNkeeUVSmFl5nFGuTx0zdptmc0TY2KgXe/+bu8//6PePL0CUVRkukcW1hC+nlDcGLLHgUFss7h2yMhCipYVRV933N7e825gnq1FfsYFZlSEonwcX6/2/Y/YdF3vVhil2VJdA6UJs9E6CsE/2RXjeCnbdey746MzjNMEnfjfeDxkydsmhWhc3zvu9/n+urAbdsy2JFcG4qy5O54pJ1GqixjXVeYModaoCwfpFfvp1FamCBNQoiR0TqGaVw8bvpx4uZw5MX1HdahOHvmAAAgAElEQVR7ShVxuztKvcdkBTGdCDFGQp7hlai6XPQU64ayzOWaT41PcJ5d29NNntvDwDA5QkACC+ZQgCh03TwvqOtaruiyoCwrsryUW2Gc0InBqAiURSF8G6PJyyL56MgMoFTEqIdbAh2F/GUkMM4YEXQHm8LGYgQ8zk0LWW5hX6afQy1PSvr/898qwPNjp+M/uhxL6iotN4adOvrunqo+SwsmGMeRWtdkxtANA5ePn7I9PZdsKJPx5rPnlFXJq5evODk9oShKmqYhoBgniRXyaeJyITAcD0w2KdyUZrPZgoJ+6ri9vWLtLCcnEq3kvXD4VTq8PlV7M1m7GOpY71E6LuEAs8UbaubSO3aHA/uu5zgMIm4InvV2Q1FW3F7t2L2451vff5+v/fBHfHRsJQhNSdKcpN1Fnm42/Ozzt6mVAq3RxYw2yFV96DuKTLjf3TQxuSQ80J7Jee4PLbf7lsGJEy4echTGB0yc8L1b2jYbxepD5eJ6W+SZ0G9TDThradueY+d4eX/k/nDETiMz7KGW8w9GxErk/v4erSBTYuORZzl5UVLVDWVVUTUryrLGeSG+xclBd0AjdGhtNGWVUZcVRVGiDIn0FfDeYW0aLKPCKENIYnF40OEKijcvtB6klfN/f3zQi//Iv18WgK8Pi1qQpnnZ1h13FMVGUtjLSohwfY91YqV+fX1FXTfkecb1/R1KaZpmBUB7bLGFlZkjF/2rWSSBniklmQiMWXJ+fkHTNAxDLzpc52iPR7RS4sSQllzTJOZVxacRhg/eEogLx0HrWV0kkZc5GVrn4j3YHdi3Pd0wLuamRmmqsmY4Dlx9fMX1qzu+f33Di/YImYZoJC92sPIAac+Hhx3P+nOeV6cEB+W6wnqbiEaGyXkG66gzQzt0y+llfeDQ9+z6jkPboyPUpeELb1zS5JqTdUM/TBAlYSUgtIosy8nLklej49UwCTyFwjqJd3HA1f7A1f1O4kKLkryoyPKCLDOcNIbzdUnfTzSFQUdZ97fjxH07sd/32GFgf38nD4rR6Ez8J8tKHoTVekNdr8RHKEC3m7gLOyCQG8hzQ1FmlFVJVZayO0ERZHEgp3Z4WCrFqF6D7h6WZzO0Bw8xP3NLJEO3YOfzplQr2S/Eed5JD19UCjt2TNNAXtYEhEPVHo+gFOcXl4QY2e93zAJ1YzJBu9KwOY4jZbPC+UCmxBUvz8U6fZomNIq62VCVFWXyqazrRiwMk5PaMPQURbEMsq/TqD9x0d93HdumQU0jRfIAt9ZKIl+RLzeiS5TjGCJNXaGtCDuqsmLVrHj53kuur+9596OXvPvqiqBF96kxeJ8431E+CE/kg9tb3ro8l2Ftpt8mhbvR0hKEEJZh0sdI2/fs+46b/YHJBb747DG/8otf5qtf/gJVVSYfSPGrDEol2z2RNe56x3/5l/8GahJZnHM+peNN7LuRu0OLNhnNphGGn494N3KZRf7zP/dH+eoXL4Xl1+RE57h/tcPkJb0u+dZ7LyiKiq9962O+9v1rPrppuTmOHA9H9sdjajHUQoeuy4qyqtmsN1R1RcwqfMw4HC1x16PigVxHqjyjyBV1WdKsGoK1qEzQIe0DuYuMMbA+PUcpxe3VKzAisSNIwkrQwmOf1/cxREye9laJC7QsfpIox05jmiEGghuxaeHXrDcUZblYlEzWcnJyxn6/oz3u2a7WQJpVTJSvhSUi0znHOCrquuT09ExcILwjMxo3yWFUliXbkxMO+wN39/eiwRg6FIqmWS8P9KcLWhsdWg3EMhCjqKZWdUMIYqdWRI8yirposMFz+egCs9+TDRMRqOqa4dDTtwPv3+357Q9f0PtIU+SUxmBQeI2onZS0TQbDsZ/I85K6zhiZqPKCLHk41nWNUpq26+mHHm0yfIwcx4H7VobNCEzWc7IWFzZvHYQMrRNnRM2DI+AV3/7BC17tWtAZNiXrDePA5CPX90KNKIpabKfblugcX/38GX/xX/k5fuathjIORBWo8pxgNHqbE52lmY784rOCel3zx77wRVT1FY7dyA+vev7rv/YNvvnBgd1gcV4IYdNkJRT6/o5rRNlflAV5UZKVDc1qTV1X+FwOi95Hbg89+tUBjSPgKY1mo2BL4LNPHvMzX/o87/3e72HvriGXDC4/WtZlTtQwRMVumtA6J3oIRUa2btBK45xH+SDYfeL4aIAY8VXJVG8p1hmDF61ulhU459lstiitaduWs7NzorNcXb0kK9562G5rw+Qczj2c0OLgIDPTZiO+o85alNJC8FOKsqqFwqA1zk44Jy7OSnVUVb1QkT9x0e+7UdbCaQgqikK86bWm0Hnq5+VrjUmpJcZQ1zVZnjGOE+1x5P3be37zhx9wmJwIowGjFLnOMQRI6qfciCuXD567tqWsT4STXRdpWytIh8ky4iCdTT+NAo22HfftwOQDUcF3Pr7if/+tb/Ps6ZvkSvBpPdvXGRnClYJoMr7+rfdwUZEp+bNdkPng0A30w0ielbI0G3tKrfkzv/CMv/QXfp6nGzhc3aCqiryo6PYdMTEXq3XD1Mv+4rDbUxQ5tYrU08jnG89/8W/8NO/dRf7a1z/ma9/f04ZMjG+HjrHvcZMk6U12lHDj/Q51JYQ1k4mGdLPZUtYNZb2iKFZkWtGGif008KEN/PC9F7z78m9wauC0LqkINEYTTY7zDu8lGUaPI3VucZPH9gp7e4XJM/aH42I1PjjPeV3LDaAN/c2Ei4r1zzzhMPRY78jziizPub+/Y7XasF6vub+/l8WSMdzv9mxPT3HeLsnrc1tirV3oDMMwMOsZYhrEjRJ+VPDimXmy3jCkttgYwzhJkmLX9T8W0PBPXPSHvkOZFcU4CflnGsmKnHyByhTWORRhsWoQCE+8Zvxg6I8jH93sOQwS2ThfoxpFkVhyyghfptSCRGgidVFRmJoxOMqqpDseUSrHa0GPRjthfWC0lm607LqR3lrKZP8wec/f+o13eXzxlH/mK19GRU+eaXQOJjhMFPvAD65uefdH76ONFmTIeUneCIFjK4zPosg5HPcURvGnv/KM/+hXv8yTlSZaR1E3tPsjGQN5VaIz8Zwce3FV9lbE0kKXCORVQdvt0QQ+t9H8x7/0hG88r/lv/vb3+P59IMsLirKmWa3EljuIp421DjeOi5/N0Lf07REZYEV1leU5VdVQ1ytW6zVjkfNRcHxgLeFg0X6iVoGLLGOTGQoCqyznFDHBFVVaZLAGGyOuKlEKSqPY5gWX65L9MHGcRupMoQ53FERMlifT1Y612RCs5Xjc47zcGGjNxZPHhADHQ8v5+bl4VsKPLZiyLFtE+H0vdjAxyLxhtMaoDbqQ97IoiiQkCWidUTUNV69ectjt6Ib+kxf95B374xHlPUWW4XwURCXLsEmMXWox47EuCbSdKHii0jRNyZuPH/GldwZe3R74zssXaGMotPhbblYb+vaIiqJfrcuS6D1vnT7is4/eoGpK3HGiyHKmBJVm2jCGwKHt6KaJQz/Qj552mPBBTJZG5xhcZHDwl3/t7/P86VOeP7oQWq0yqEyBE6uI33j324yBNKDLUkeIapI4mGeG4C1+mvjJt7f86a8+R8WJ25uJMDryPEObnMN+T6M02oM79sToKXKDHSfq7Vrs9pIPo46KruvJihLrIk9Mx3/wT3+e//Sv/hZ3vSJTRuZTrRMfRYbd7PR02QxP44SdZg93h7Xi+HzYHwRh0XL7FmXJqmnYrNe89fhSgu4OAx9aj7I9yg5k3tF4x5pA9NJuxSR+H8eBi6ahLBTeOpo8x1pPXeTspxFiIDN52pUExrEnM2LC1LY76nqFip7oPdvNKUShU1xcXCzGrdbahbc/m3VZaynyghgiRdrO9n3HZC1VU7PZbPDe0bUdt31HBG6urtjdXzNa+8mLvut7dFVjQ3IZ0Iq2HfDZxOZkK8udKNRUYiD62d5PYiIrU/L4bMOXnj5l+lyLiY737u7RWhYKwyjFkZi3aCLndcNX3/oM51klFIIY8HZaNpNKS3JIANp+pB0tbW+TVV9gP/Qopekmx+CgGwf+x//tH/Af/sv/AnWhRUKvpPDaoefd9z4gKr2YrcoN4mgTbFaVht19i46Bn/7MJVn02GFi8oGuHWmaCqVgdXaB90EMkJwsze52B4zOaMcd9UokdEMM2OTFftjd4zx0w4idVFrN51TrVTI+lZ97GkemoSMS0SZPHo4F1WpD1axRWuMmi3MTcRqxQw9auOuunwhx4PF24j/51S/x6LTm//zmNV//QccPXmnc+lTS+o57vHKcq4K192zKQmgYsRYo13v204BCpIeddeTlSmKK8kqsN+zANI1YO8nWdOyZ+mEJW2iaNavVirY9Lv1313Ws12uqqnpQxAX5+tn8am5/+r7n2F4RoqepG6KPdMcDbhqTT5EQF6dPU/TOB45dn3Bg8WlfVxWPL8+lFbAeskKCu5JXeZ7LtzRKU5icTEUumxV/4PIpDBPX+29yM0poWz+IjZ91luA9o7V86dklj/Ia4wLT1KG0tENllrNvj7jJkmU5zgf6yXLsewYHRLkpFNBPnn7yaZun+a3v/IC/+413+ZM/97MYH0AFsgy++cP3uWp7AsJRGccB6ya60dEPljx5NPaTRRNpjKR6qyh02Dwr6NqRsR+xVtonFTx2mig3Gy6ebsi0obvfoWNAI8jPOI6MdsB6x/X1kfsBvn1zxDrh23TtkbIspbCrGrSi71r6vicExzQ67NjK9jF5cBplOK0Nf+lf+8PUeuTRow31umbYtVyuc9bakauJWgf+/M9W/Nmff8TXv73nr/xfP+LbraM4PaedJu729zzNNbXWGBzbTc3h0CUHAi2emTGgdC4J8SpQFBmyZW7wXoQkh70opJwX5KWqGm6ur6nXguJcXV3x9OnTRRs7i8xnsfxs8Ps6x19o45boHFe7e4ahW6BMQ4abpmTD+CkIZ4d25GTTUJQlvZ24Ocj13A8jTVWLGecky6HRWzFxQlbTucnIVcbkenKjeXpxTrs/8pmTc159/AGT8kSUnAzOY1TkfHXK0+0pTVHIU9wfiau4LGF0hGlyDNaz6wQL7yfHMLnEZfE8vrjgtr3DR3DekZExasX/9Pd+gz/w2Wc8f3SBdZaum/jN7/yA4BPMRUwb5Ejb9dJb5jl9OqlyrbGj4+72yGlWsylrum5PWdc8euOCYRrJywpvR7RYBON6mQ0O+z1KKc7OT8UhLlHONttTeqv55vsf4Z3mjbOa7111+KAJzkKueVJHjoNj7BNF2YlU7nUxOUqoCr33tC/e5yeeN3zu4gRTePZ9T3d3zyHCoyen4jYdA8YPfOWx5yv/9k/za9+84r/7m+9C84jV07d4cXPFvh14Xkaq0Yr7Q6bJcyFz2ZTAWBnxGp2dyLyX8GJjMqqq5uXLjxcSnUaKc+o7VFp4Ho9Hzs/PsSnmfl4azqov2bQKamftxJQ+5yzLxCJwoRJHnBsTfv8pqcXWBrrREu4PrKqCarNicnC7P4jbWZalMBd5GvM8py5rNHpRH3nEhqPOc843G55vT/josOd6HOSB8Z51UfL5y0uebU7YFhXb7Sn37YFhGlGrfHkTlFZkKhOk5tjRT5bROml7DDw9PeXR2RnffXErr0srbPTk0XB9aPkrv/Z3+ff+7K9QZIbvf/yKl/f3gGwwvZcBsR9Ghq6nKHJRCSUar3WewQVGF7i97ygV2ClgtOfF/QvqqqTbHSmbkrE90sTIyekJ9WrFozcuUNowDSOH/YEmM/joGaYRj+HZ2xc8xfDhbcc/+6XP83Hv+Mb3X/EvfvWz/MIXH9GGjP/2b36Db73oMHlO4giLVeAsKDcGU2Rc7Qdud3BxL546d8eJl9c9jx+dgDH4MFGUJdPgiFmGO97xz3+x5A8/+yp/9R98yP/6/1yxefSU3dULvjvIsulzm5oYFS/vjxijOKsbRueJJkOHKBaNSgyXJGlkYJxEgO6co+8HPrr7iDffehuT51ycnDKOE13XcX19zXa7pSgK7u7uRFxubWK4grUjC7MPT4xe6rLvGZPIvWka9vt9osfoT+dlWZYFh3agG5yQgELk0LY4J+qXyXry3C3DVWly0X8GT54VTHbCq8jkJmwvQ99ltebLJ49og+PoLWPwPNmc8uzkjI0xbIuCLM/o3cDgB1YqIyYnqxAVk3eMznLoO9rRYl0g01CXOauywNpJHlQ12+Gl36vgN7/zAz7/td/ln/rKl/nBxy9TbI1Y3Y3TgIvQjRbvA5tNw83tffr9gSLRbldVzummIq8qjsee/YsbslwxDBM3tweyHL7wxWe888V3hKNvNDF4bq+uKPKC8/Mzhq6jXq+ZvKNay7LnrTcf81Nvn7GqM+qTNbfHiVWeMw09GMN/9ue/wjc/6Pjr33jBd17IKj46nVoDuf53x5Hfee+en//CF5iCYmxH+qB48vwN1pWhbQ+0x4mLR5dYFOu3H7P78CV2GnjzdMu//sef8cMXd/zGj97jyRtvsL91/HC0wIQaRu76Huc971wYirTgi8FT5AUh+MXTVHYpig8++ID1as16vaZtj+x2O6q6YbsdKUvp4e/v7whBgh2qqqJtWzIjg36ei1tzd+zw3lMUBeM4iprMmOW0v7m5oW3bZL0o9I1PXPRVmWGDI88yjl1LmSu2dY2yimM3cLPbEeMGoqcsCpQ28rTPK3znmOxALCK+HVlvGp5cnNG3HUVRSpJ0ZogEmjzntGpQMXLse267I4OzlNahy1zIZdPEfuhpx4kQFN6JeWmV5Zw0NVVh6JzE8ARC8mxMxq4hMgH/8z/4GpvTLe9f3eKD8OWDn3BOMk/HYVyE7s7J3BK8pyxynp9vuGxyVAzs7g5sNqecbE8oaoPBcXHecPn4kqIucdaTG8Mw9eg8JytqgvPc39xSVIKoVCqw3W54dL7lsLtnlYutSq7g8WmJtR6lc+EP2SNffVbxU29+gb/893/A//HuS0KeC00gOLH9CIHfu+q5biOfLUusdazPtsJqzOBk9Qimnv3NnvXJCYcPPqbdHTh9ck62qvE3d/ziH3jC3/z6b+Bj5OLshF2744WNfLYq+fyqYppGCqMk7lIBwVGWOV13XJIHrXMc9gc++OB9xmngSz/5U2itqauSGAO3N1c0jZi7lkVO17XM9h3ee4b2iLWW9nBIVuGeYRjpOpbQZDF2rRMBz2CTQa6axmVX8omK/tAesdZBDKybhv2xJYbImCkiAY/QAE42jVxVx2vKPF+wcgAMmFVOf91iTM66qVhVBUVZUOYVZyengkzEwKpuUpEG7todKhOFlJ0munFg8pZ+Gtm1PYdBoLrcGC5P1rx5vuXp2RnvfnwjksC0YidjIS8FFHfdwK997RtcnJ6IzDQGhr5j0xRURcZHr25YrVYMfbeI0FUM/IkvvcFXnm2pNOR5waFvGW52bDc1bX/Eu5GL84349YSIsxNGZQzHjmYj2HW37x5sDr3FThMBnVLHC5wfePHBx2zPtjx6epE6f4EPy6qgKgrCfcu/+cfepjGBv/5bP6JrxyWJIwuBP/dHP8NnntRioa5lCTjvUdpjT53DOE7o/TGFaYCdTaFy+MxbJ1xuKq6ub1hXJdrkvBo6PrPa8Ma2wE057egYVCAUOcFLnz27QkvmU87bz54zThMffvh+eh8tN7fXnJ2KkGQaR05OTkRqaIVH45xjtVpJC9M6MiU5CNM0pm4i4sIghDig6ySfKjO5iJCMliTIT2MBMpt82mliMOIpeOxHqHNK58n6AR8842Tpu6Ow3TLDmPjYZVEI8y0GWGfs7w744Hl0fs44Wso6p2oqVlUFSZHfrBp+dPOC+6kXwcFk8UQG5xitY3ISdOYm2eKuywKc46xe0ZTp6yOomIhXLi5uBERhIhZVhYuieyU4nLdc3bbc71vKPKMsDO1BlF8zcnVS5ux2B0ZtUGagyDMIirvDEa3FB0cpcYEY+pHhuEMFCU8Y+4msyOj7idvdkU0/su071ps1eVHhp8QIVYqnTx9xv9vhnKPZbPDO0R5aCqMgOowJbAv48199xs89P+Fbr0b+9u/+iB9eH8mM4uK8Yb2pGPt+UTNN3uInhZt6sm1NUZXi+ra7RyuD8x4VI9PkUSrn7PSE6+6e+92Oy8szemP4aLBcak+0lrLMseOInxzj7h4Q3tQwjEnLGjl2A88++wUeP3mD/X6P4oYXLz6irhqKUk703W5HUTdAQs7shPeWplljrcd7S1FUNHGDmywKzTjOG9e40KVDCOQmYxgHbJJEfuKivzw75/rmlhgjfd9TVw1DGMk0KNUSQ8Wru1su4wYi1HkpJ6pSeOfJcwkXDkFjC+j9SHSR9WbDyUWBdZ6ooGgaVBR53ovbl3z71ft0esK7SGHk1mjHgXayDKNlGB2gWJUFZZbx9OKM7apBac04ucVhbDY4NYj3e0Dw3boqyY0mOEc3tEzjyKET191f+tm32XcDr14GYhAKdZFpTreCx08+SJxPCFjv8JMjN0KrjXgO7Z7MGFbbijeePSb6wN2rO6pccTgc+d53P+Cdn/gs2/NzusGhhkPiHhmikwAyPNy+uMUODhID0U6Wdr/DxUjuPMFGHhUTn/2D5/yhZ2v+l996n3/4e9d898N7fuatNSFaytILMzZG6qqgqUq6rmO1XnO23S4inNzkRBRtN+Imx9uP1rx3fWQYhlRgipe95TrLWXsRelR1ibGBgphE3RKJM40TZVWz2mzY7Q9stqfUzUZIY7AEOsx0g8l78lLaHqLAuXZyNKsVStVAUogltq9P7gkamRWDl6Vi13cMXUtMplafuOh/6We/zN/52m9zsxPIzaQg4KAUU4gcu5HcyAmfa7HRw0tcTd00hBg53B+wUyBrKuJ5STfs6TtLaCXdo6pkETTZicN45EeHK+6iGDIFqwhiy8joArtjR9tb2sGKk1aekWWGvreEpKQ6dD0itUumP1EiH7UWa6r1qpI3WEmv6KzDIwS1yQZ+93sv0ZrFXiRXkZ94suVnPrPip9855b0f3LHf97Q+0KxqQXXGiWADhy6gjOVzX7zkyRvn5KUYzTarR3T3B55/5gnlekvU0gIdu5bgowigE3fk+uqaqlnjVORwvKLZ1pxdnJAXa3RWEkXrxzQO5Pkaa3v8YccffFxTqMe898E9t184pyoljNA5i1KKaehQq5JcB/Z3d+xubohKk6Vh0duJ7arAac2bpwVlodkfx2Uv0o4TRwxnhZD/vPU0ZUYePUWZY5WEM7St+OKcZJecnp7SdR0Abzx7zu3dTeLMh0RhNoTosX0L2iSVWQEqcNjfs95s5bWPAybLsP8vae/RbGt2n/f9VnrjjmefcGMHdAYBkCABSiCpUqBQklmyNaA9tSYeucqfwp/BH8ATV3lgD+wqly2VqZJNmjRtCgIIshE63O6bT9rxzSt4sN59mpzQVeie3eq+fe7de73r/YfneX4W9BjY2vdx+tN3fdwZuSGOOImwi1/50H/w5gPWmzU//uwJ+0NDU9cUeY5ShmGIDNJtI0gqQ2E0VV1jnSPPCvpDoK+aO0+m23fU/cBBtvSuozk02CGmH/eDJSjosAQlGEY6nwgyGii859B2VO1A3Q5Y62NKLpJEKhKjR6atxrmjM+gYs3cULnkSrThZzEYbZJz9Hsuzvh/iYk3nXF2vCQROJ5J/8Rvv8sPv3OPxacLpPGP1nYd0NtDUlqYbWK8l1zd7ai/Z7mvu3z/n4t4CUxhECAQ70G92uN6SzzPeOY8+gWGI2p4XT68psz2TZYlOczqZ03WWZZmz2TQcWofSCuFDVCS6mDqw3mypDw1N23PbCA6157TQ9AfP5fbAB++cI6UnHRu9rjqgRutiTCxzaBOlwkEZhBV4L2CQaBFX/1K0VFXFyXKB6wS7wZMWGcIPpErw8HxCnwaU9ExnU4zSHPb7CDne78jyiOLcHw50Xct8dUqepew2m5i8bPQoE49MAy8E3RgnKJXg+vI5YvQODH17d6iV0mgdI2mUViRZStM02MHeWSJ/5UN/frLizYtz1vsDT9wl233Noa6ZiBmHqmJWpKRJQl33ZLOUpmnxPlDXO7qmJUk0s2KCHSy1Hbg5VAwEWj9w3e2wzpGahN4PTIsS23rc4DlS+JyH3lr6wbGrOurGMow5kVpqssSwmkx58/yUIjEcnKPuorNJHDXiAY5hR0Wakd0JlaK+A6DrI0BiNptyfbNjcAEVPD/86AH/7MM5i8QyKXOs6zAmYzEtOVkpvO1xjyZ0/Tl169CJ4fSkiP2JtfjB0t+uI9K9nKATBcLT9Q7rBavVkp/89Zc8eT2w2OyZL0o2Tcf+0LGte4pJye12z8nphH6/wbuATjOUmaBNwmJVYJoGux/o5MD1vqftLa9uG+6vDkwmKUWeImQYk4NrzHLC0EcRX0w51jz98hIRFI7ATeXIdayZtTHUTcPpyQqk4Kq11POUtB8IWrHveibpGNokDUL2ZEVOVVUMXYN3PVoZ8jxnv1vTNi1KafJRcxUhDl8hRiG67Xzv7nQ4XW8ZrI/lk440des9bdMgxjc6MsKo/1bS8q966PuuY7WYY4IgUYppUVC3PXXTEAQM3tP0PUelX9e24ANGaax3KCvY7mNp1LlInG5tYL07UI/UiqYbOIaM5mlyt91DELeqQxhLDxu9rNahtKRINJk2JFKwKHMUnkM30I2ACCHEuJk7ZrYEFovZyCXV2PHmcAGGwUXqBjEuTkl487Tgw/OCLFHkRUpepCOLSrHfbimLLG5FQyDLEiZTGTUqwSO0jpZuHVWh2XRG0II0S2jrfvyyNTYE7j++4N/88V8yrVImm5ZNbbmtOiSB1XzC2aLEpDlh6LDdgNIJVd1QFLEZnKU5Ojc4KqwP5EXGpnFsqw6lY5OXZyl5kSN8jw+WrIhI1K7t6LoBj6RpWtIiJwSHH2EJSZJQjfk1SivafmA7BFYhoFHs9o6ksWTO3ukqRfkAACAASURBVHFpi7wc00SiX8G76E/QJuX+/QWT6Yy+79glmu3t7Wi8d3dJw0fFpUCiNeRpitEjOC9Ea6pRCqUmsSwbHWF93+H6fnR5/d2H/u+c7aR5Rp7mTLOcd+8/oEiiybkbYilggU3Vcbtr2B6q6EbSiiE4Bu/p3YgvJ9bVBEHb9LGBRdAOlm5w9IOjbht21YG270ddfhxdHn/tgsP6AaUl8yxllqXcmy94uFyRCEWa5uwOVbwVjkFIHLN7IDWG5Wx6dIRGN7+3DNZinaXMU5qmBgLvnOX8Fz/8iO9+cMF8VpDlKS+eXnH56oa+65mUBVoEwjDEOPChx6RRvRnHpAqpFfXtnu1NRX3o6aoW2/coIZhOco5K8cf35vzTH3wLhGDfefoQfdrbzrKvO+6dz/DegUrIprMoMR4s/dDR2C5OymRgtUxYLTKMiiI7G2CwATMGKm3XFS5AUJJ0WpJMC1SSMAyOosyYLyYEAkrC/dUc+j7Go3hH3dTRgxo8L+oOTIIjXgCXL67pm45uVEtKFWEcbdegtIl9g9b0fUzFiHF9UXs/mS8QIo44IwDa4wZ7l0rc9x3VoUIimGQ5mdHRMaYkOBuFiPYIc5DR3ywV2vzdy6m/89ALpTg9XfDN994mM4ZJkZLlBqElzTCwPTQ0vcUS2LcDlkjF2zcdu66jcgP9yCSVUpIYNb564rZ0GGLzobWKBLwkiRHcQDcMDM7R24HtoYr6HCkoM0NiFLkxnM5KlrMIHXBjrOCR1QTcpR4IYDGdk5gUrZPYbI/ThKpp8aM+u227UX4sefb0NTebA3Vn2e1q2j6g8hkuxIxKkLS7mqHuKZdTUCJOEqzHt45+WxOcZ3G+IikLDvuel1++pDnU7Nd7XGfp2g7vLA/Pp3z/O98gyWScdgUos4xpmbKcFlS7ivV6z6HqaDqHDzISvLVEaYHSMJvkJIpoNg8yuphcGKnfCqlFVF0GT1BxS73f1xFLmRims5Llcsp0lpMkEiEsUkUTet00MX5QKvaDoxOxxLjZbFEiEGzLEZsTQiAxmu1+y5fPv2C93dA0Eay33+9p2iY+XNqQ5iXldBFjWCBGlwii/GNMX/Desj/sqOs67jXukuWOQVjx11prhIy4Hv11nFNZZhh6eOPxfUQQvPXwHr949oK//OxLXq63eA9VV8XD2rTMJiW9i8IfREzF7YaeQXzFWBUyHsZ2GLDOYbQhSQxGfpVWOzhH3bSYxMR4vt4BilRLyiQlk5KLxZzFfEKeRWNy3/esFnMSrUbvbAx3IsQFzdnJHKNi0Gxft3dhsUNvx+TdqOLDBe4togS2GyRtH5O9JosFJpEURYJrO7qmIZlPUMbQVS1JphFITBYN0EPdstvuWZyexCQDJLtthXWBLJsyWSxorm6RAWSiuH824eTkm/zRn/4VeWJ4eb0nv5jz5bMr8iy9225DIEkML15ccXGxRAgXozm8Q+LIUsXL24qXV3sWucQowbRM0OKY9yioqwY7eNqmZzKdxBQ5LbDO0w2O5zc1VT9Qhpgs3fVdFI0pQR8cUmoSET3Hl7e3zJsKM/+qpNQ6NvHCO/abNb6cRDuhGidmzqLTBCMNYQRBSOISMoRIFTm+NaSQKBGFiVKqr8ofKQlSRLWnVGMUoPjbkYm/yqEXIsbpmdTw4OKUzc2OcN9zs9nw8mbNMGY7tl2HErA9NDjbRudTlhPTWOI/g7UMdsAkCtkFMhMnQFortIxThbs6XMAwKgqj+30gTzSlSVgUBefTCfeXS+ZlDPbsXYxtnhYpiZJUQ7zdhYw1X56mpGkS9eghMPQ9LgSawdLaIT6s7YBzASXgZlvx1180tG1P0ySczkokgjfff4RWMDSBZFJQPjzDNy2hjXHh3llEkNiRbzV0Lc4N1FVDb3uENiA05bRkv9ni+giVyFJF37fMJzN++PfeQ6B4sT7wH37+nHUt2dUDaaKpOjeatluksORVi1aSNEvY7Wq880wSg3WCL18f8IsIXCYEEg1ZVsTmXQiapkcoQdc25LpEGY10ARkkkzSJLKveYZKEum6ihQ+oB8cgJX3f0ttA5wLV+obFxZsc6fJRBVlye3sTJ3ldw/5wYH6yIs/LMSRWEJRAGk0aCpx1DINDqfHAhpjWZsNXeaVwNNYcI0PiMlDqWEKHsRF24WuMLO9i44iMpaJIua9P+P6H79MNA19eb3h1s4npZ8PADVuU8EyK7A5xqYWkSFN8PzAEQZCCrMhwVRddR1KMoytF08fmtu16BufQxA9nmhkenS6ZJilTk7GcTJiWBSYZ0Ys2jjVnZclqNmHT7b6a3EjB2fkKncSgp67vGVyPC566iZOeNE25vFozeI8BfBC8vG2xDg5NxrPnN/zhP/8ewvaIJCGZ5TEmrOtg3KZ2bQdCMnQ1+/WGrulRRmGSBGUGktQgRMwCVVpQ1xW7XcVitWTwns4KmrZjPiu4vtoyTyUn04J//eefc3Ey4Tc/eECZwNOn19xuKh4/OuXevQTrBvpDi3OCvndM01hH//LVjlTPqPo1384vwHv2u4bUgB8CwUZXmElSnAu0rWO377jZH1hMMt4+n/NkM4zhrHFpVBQFQ9ex7S2zxOCEJTcS5bpxNPxVoNR0Osc5y/r2Fu/iXmFze408gTSf0LcdJkvRo2Yrl4qu6/HOjgC2rxKInXORXh8AIUYh2hEioZFOjjuZMXFZml/90B8dLEob0iwgkMi6Y1XO+PajR0gEfT/gEHTesatrskRTAnUbCXhCQ913WOsYQozYYPA4fMx8DDE71MvoD22PcmEZ6/5VWfLWxRlFohHWk5uEMs/I8+yuHIpueIGWkvvLKU+ut9jxw0+NYTGb3IEVuqaK9jrnaLtuxK1Lmr5nCIG2d+xax1lueH69p+9b/uB33+fsJKc91BgCZlnGKMOmj04iKRgGR33YMwyWobUkeUpe5qyvb9nuKmbLJSbJ8EPP5uaWzGi2PpDmKVebioBApwnrXcXnL66YliW//c2HiCTnL/76CXkSeHRa8uEb73O9qXDSszrJuLzqaJvYX21qx7axONuzrjqy6RTbVbx8veXxvck41IiLJWOiLBwRt8mmiICy2XxB23iK7Cscpxj7LyklTgg+31asVjnCO7qho20qJmNW5jFXUiuNkhozmkTkIOmahsNuS9e1mCRjkZyTZhkYKMYIx5tXz5FuwFqHH+JBl1IRFHfpZUeDSZJFSIhzA95b7NDTNe1d1PqvdOi/okToiEZJAsEFyjzn/ukpiTZM0oyffvmcJgT6wca4DcAoSZ6mdMIxzVN2dUUQCkGgqupIH1QyzjCkoPeOECTd4EBI/NDz4HzFW+dnzMscJQRhVFSWRR65RUcwxFivKik5m83R4jluXFNMJxFoLJTCW8swtKOsosdaT1mmXF/f0ttxdu88n93uSe/NSI3hh7/zAT/8wXto27He7jgpC2ScnSIQMXulqVEmIy2nTNOEX3z8BW+ezaj3e9rO0bYDUynI8ozBpty+eE3oWvLMRMONibomkwhsL3lwvkAJRZYq/t6HJ8yTjmmqUDIugr710X2ECNGHnBk2u57+yyvqYcD5gUyn+DBwtamZJ5LnVzvA8fB0yr3VBCHHhrG15JMZXR+/19QYFqXkuq451DFCUYwU9ciLjWVFZT27dmCuBYnRTBcnOOvuYBliTKxI0pRnz55xdnbGyWrFs6dPqesDoT4gxxS1xek5eRFlxcc5+359TV3tMUlE7wwjX/YYEwIRet3vtpgkQRgTqxGpSbP86zmnYiyDiZpwEXPclTFkRc7cTtHSYJThbDrjalfx5dUNzzdrvBdcHw7kmaVMEkyi6I/gBAdeimjcIMLNutESNthA1/WUecY0L3jz7ITlrIyrae+RBnJtYsqXGcEQPhLJjdYYo1lMJmTG0LoegWC1XMTtopR0XQ8+4IKP+YljjF7btgQ33g5ScF11/OjLNd9YJnzrvQcI39EPLfk0oesbMpvF+JGqpWksppzQtT2pNvRty8MHc7qqYb+vsRb6zkd9Co5ikqMen5Frycunr7m9WbM/dMzmE2w/sL6+Ic+nCKVou4HtdkdCoO8aXK5oW0HbdGR5QpIlPFzMyZMN00yybcB88gKhFF+sD/z86Q3fe+eMfeXp+h2bTcUwnBGGgSxP2B46lmtJkUoun18zmUzRRcHQ9PSDG1PHzMh9irGCkXHlyTPDSSbJZwtWj95gL+KF5RBR60SMjFksFuz3e05WK5I0AhfM+H3sNzcMfc98eYqdTlHaxNAmEef7w9DH89HGXKBjlOzxUB99HCJYnHOU0ylJWeLc16ALeidGlpRGymHkngpSk2DTEucEE+sAzWKyYDmdkTzXfPLiJW3bxYNtLYgRiTMCwcbTRdf3CBE78X6wOBtp3/M84+2LUxaTCalJxqD+KI9NlbnLLoSYMOCci1mUQjMvC+ZFzr4bMKlmNs3HSN8ooY1RHIG+H6JLp+sZhjG2Toi4UAnQDAMPzs7ibN0U1Ls929sdy7MUOww4a9ntGhAa38XXrvOOF1+8YLPexgOjI150Ol1y9fyGR48XNFevsV6wGTybzT5m1hDo91smp2eYpOTZ02vefOcR/dBTNQ6hDVmSUEwmzKYlddOx3RyYTHKm8wHf9yRK8WiV8fj0fSzwD79f8Sd/+YLt7Y5D53BOspomXF7vuH92EhdeTeDnT57x9v0ziiSlHyyXVxuev97H5LMQRkqhwI59k9bRi1oNjkF5vDSY6QIxHEV+0VudZTkiE7z3wYdcvnqF0glFMYPgSdIEAWgdqA5bhr6jaU9YnJySZSl5ViJX97i9vaGyB9KiHMeX/s58f0yZFjKawX3wNE3N4BxZVvzqhz7mIEbEqVIay0BkTMX6MzsmFktD1dRMh5R3792P0gHrcD4QnKfrh/HVFFGPEb8+QsMIcbYsBMv5hPuLKefzGcvZjKIoSEbQViDWdEEeZ/3cTXuklGip0EJSasOyzHi52TKfTsa0qxgN52xPQNC0fTyUiWF9ux3XRIG/Scz+xsmE/+pffpfzec5uu2e33nP+8A2UEgRrsU2L7XqSzFBv90znU7pDzeHQ4UiRicAGz+3W0dqG89MpoBm8wntxp/ysqwZtMg5Vw6F6hhCa6ayg7VrqzvHi9kCep0iV8Oz5JeenK7JUo2VKXVm8ryknOWBRwjF0LUlW8O5ZysPffYPPXm75yec7/v1fvaDMO6QwTOqOyWzC/fMTtlXDrm3Jihm/eHZFHRSfvq642fexzPTuDnZmrSXLMnqleNE4ztKUi4vHiDQHa2NVIGPtn+UZWpf0/RRl8gho1mtOT0+iP+KwxxhDWZY0dcNhfY0IHnV6TprlzJYr0rzg6vqS/X6H81/pbpJMRzXvuFgEgTbqDgBY118j9yZqHhwiaLyQoFPA4oLFSYc0CjlIsjxlCJ50GFgKwYcPHrIsplRtQzO0NIOLs15r41rZaILzuBEEsJyWlCblYrUkT9NY5ybmriwJIRJCIu4yCsu0UndRD0YozLiYUEiWxQSjrjlZzEmSDCkFfd+Os25L13WYJL4trPN3poNhJBWe5IZ/8VuPKUVNc/D0Vc2DRw9I8wxwiGCjeUErpIQsTejbuMCZTmfobqBqOhaLOXneU04yNte3TCcPSKcnbDdrkIbJfBUNz0NPY+Hp8xuUl3zzm2/hgmNdt/z8+ZbCaAotKRJJEIbTZUmiBJOyAKmo64bESPa3a7RRKK3YXW1Aan7t7RNW8wQtPc9e3rKpLVpXMRuyiBMtVeT8/MWaL286Xu8Hnm169p1FmARnLSbP7ubnkaYSuO0tm17xaDbHEv2xhJiVH0Lg8vKSsixZLJacnZ5zfX1NCAJQ5MUEZ6PRJksLtE7o2obr18857DdcPHjEYnlKWhTcf/QGxWbLZnODtxapFZPZFGPGZeIY6EoYAR3eouXXQGq2fZTgmqAiukW4GH6q4pp9sHEaI70nyQy5y1EmLoCyNOfQ1Oyb6NA5VE00KQQfYytEzFw3xnC+XJIZgxmR6jF276v03L7r7w66lDoG/8chfETiyBjRJwIYIZkWOfNZzH0MweN8oOuauy+u7wbSIqOu60hTUYCI2YoPFhn/+e9+yLfupRy2FVjL40enHA57hPDgGgSMdWbse6azKU1TUe0OHKqaV5c7mrbn4aNzdBIXRz5EGmJnB5TJORz6Mb5iwmC37OqBj1/seHwyo5iWXG623GxbtoeB5YMFgh6RCDaHhvm0ZLdv+PL5DctZRltVZEnCrJSkeQJSkE8SsjynrrasL6/4zXenvP/GnD/6019iN5LbakNWZnzycotjw6tNQ+clm7rnurEMHtIQ9yv5uFx0Y6SjkBLrBaaYk80v2B5q/JiUHIKnKEoOhwOvXr3k9jaavSeTCYvFEiljDMp8uWK/u43CMSlRSUIGtE3Nk09+zslqw8XDx5TzOef3H3JyesFhv8V7GxOjE0NWOIIPUZczmpCkFPivQwy31kMS9TtujJeLkK+jsk7FgkdEyG05kcimjc1KEseKk66gamumaYbtY3MSubRjSm+WkmU5Wusx/lrfhXAeA56UViirSJOUIklJdFy4aCGRxIdHSxlLJSk5nU15dP8CjI4pw0M/unLiPiEESJOE3e5AJAMGrPdMjOA/+e5bnGUBFEituF3vePj4lCxLaA4HkkTRVD3dACbNaOqW6XRCYhJebl8hpOKz5685Xy3x3tJVDV3TkeQFg/M0TUdelOjEMHSOFy+ukUqzPgjefnDGd96/x/XNmp998or52YoH59M4BkxgsZxQHxq+eHFDlo9oySxlYgzWDhSrExIlyVJNlmfRyqcl3/3WWyTlFC8Snr7a8+9//pJyUvJXn2/4+esDlfUcmhghEgjIUS8TN6CBEGR0WI35NEpp2sGyUxlidkrbDjEvf2xSj6PFJBmN+nW8dI4BvMcUszS94ObmGucGrJWj0nVG0zSsb16zWV9xenaf03uPKKdLTk5Ox3MZZe1BBbwIJM7FrbePzC33dQ69s9FYkQwjwtLaaLEbx4GR/qDi+MxHPpNODCgNxsYmeERG2tTh7EDfd/SjIydOBgzGaNI0Q414xr+JfgnECHBpHJkUMVB0/GCjvlOglSI1CUppUmNZTScUVUV3nDF3XTzwbqDu4yIFccStE1lUwfGtR0u+sZpgh5qXrw4s370HKF49v+H+2Wx04htc8Ji04BefXTGbJJRFzW53oK4tN7seZ2E+K6gPdfQMyDhWbLot00ketfuppq56Xt8cuK0df/zT5/zg24+YTnOurnc8fniGUJJH95Z88tlLtE54cbtDA+VsQppoyknCbFnge0vbQFrOmMwm2MOazc2Oan8gzw2TSYboG8pZyh/+8EN++Dtv42zgqlH8L3/2hP/7l6/49NWBpu6iylJAECLaDV2UZygVHyxrHVoqGgbarEAXE2TTsbldk0+m6BDzKLXWpGmkjRwOhzH5LF56xpjxO9acn9/n1eVziiyjpMSPjWjTVDR1zasXT3n9+gXFdMZ0Mmc+P0GnKWmajsE/ERQRMVA69px8jfJGShkhCKNk07m4ag8iznVdCPHVZPSdeF8IidGgVLR0GSkwStAPAyEYuk7TjNQIpY41eoQ+JEajtLn72ccbIQw2Uvmcj3Vdkoy4S0vAUaQJkzwFEV/BfRVouwGRJgTvsEM3Lk5iM5ak8WeEMcrPe88y0/z6Gyu0sPQusK8sT57vSLVH64rVqsQFz8tX10wnMxIjOD2bkmrBerOjqmp2h5Ynn73gOx+9yck8Y3O7ZnF6ihCKytV0naMsYH27xSE5HFq81Pw/v3hGa+PYdrNrQUr6vkYnGettzbpuOVlNSRPFdncgSzqyTJOOb0pvNFmW4LqKeu+YzU9wdo1pHdtdjUx6slKgS4tm4GSqqXcd2lj+1T94gz/87Yf8v19U/Nf//f/FuiFyAUayopBRnKZ1xNwcozjoBW2Iy8ksTeiHDmNzjInLqO12S5ZlAHcHv22bcQsvI5pIa5Ik5eH9N7hdXyPG2b5RmmAH2qZht99x2O+oDwcOm1teP3+CMoY0zUmyApPlJGlOOlYLSkmGrv8ah36s3fs7dmfETvoQBV7e+zjqC1FsZMYD60McUVrl7g61ukuk1Wil6fruLqE2rq+jgVuKeAubJMHoKN5yWjOMPyf4Y8pXBKylSUqZFSOdA/Ks4KZ6DlKhpMHZdkSqx1kvIQIBhmEYkTZgneNiNmGWa/p24LDbodOET59ecX81IZWCJ08uWcwLXr46kDyeMFkaTleSvu3ZNDXXV2tub/c8erRktcpIjGF+9gZ9Z3HDwHySMbQWiWRfOZ6+vooPX5IzeJgXCQjNs9dbjNE0dY+wkh99fsXNtiVNt3zj8Yr5fMruUDHPErpm4OpygzpSzPEsjEblBSqp0abBpDm7fU/T9Ow3e9Iiw7mANil959h3HSrRfO/Ngn/1jz7iv/nXP8WP1r1IDYmEdm0MdHE5ZIzBB8/17Zq2a5kkU+azKZvNmvN7DwlZRnXY09X7+FmP/KuLiwustVR1xTCmOadpwBjNvXsPI3BaiKiaFCXT2YLpbEnd1rT1gbZu4uU3kky8c1TbDZXakaUFWZphnaP+Wod+PLDBRwnnMKZV9X0XI7XHmyCWOXFjesQyHn9/ZJhGYZAa/Y1HDmuUFn+1PbPDgCBmqnBM7hrHkV4IJBEJo8dID4UgVTpCG0a+aNV1vN7FJLAgAsMIZzvSMQASk7CpNhwhaQTPsoi5m0WmeLA8jzp/67hcH1hMC3708UvOlzPyTNMD3g8RFZoknKzmCOF5+91H6CShqSvyoqSYFFw3txEQVlUcKkua5VztK2ZnK54+fUkQFiUURW64vNngwwxCz1U18JOnr3l+U1EmisvDgHm9ZTXNydKUXd3ivSXXsJxllFnKfDElTTMON9dgh7G38lzfbjhdzbHe02zqMXS3YrY6Beto2p5haHnv3owy0dTuK24vUt31ckrJEawWv/v9bjsmJvcRftHHkNW73zsumZzztG1HURYYEWO9j2nFbrQHRv6Axo9pzahYQZQy6pfyoozgv9EOGsZwKYjAvbqq4py+H+ia+lc/9D4EhGd8zcVFRT9Yut7SNk2kZIxaeDliWGDEtwhxB+wKUqKI6BwjFV6bqLpUPU3bRvQO49jLRLpeGOOdI5bGELQh2GNOuQQxLquSlHwEfQXgan9DPVhkmuOCZejG0sZZejvc8VGPX14ExwmmueK0MMyTQJkaUq1QwrP3kl+82PHyqqEaNB88XrLbNjz78jX3HpxxehJ9oQJJnhckRcJkWtI1Hb/82edcXe9YLeeURYo2KVfrA05oPv38FUmS0vaOTTdAIjjROZ119EHwJ798xZPbNi57kJiDpe72XG1q7i8LFllEeRZ5RhAKh4qy4aYhNTndNnpVpZLMpiXrTUVV9ZycLLm5usWkigdv5bQeDq/WVF3Pg4tzPny45MdPdyhiz8OYpvA36X0ARmn6tuXpk0/5xptvMNESWxoOuzX5dB6nc0LG0Fkh6YeWq8tXnJ3dJ02zO/P28fDHfUBCmqUxdtB5inE8LaREmJFlPI6uBYy+2IGqqakPe5q6om9jyO2vfOijZS8iDV0I2PHwuNHZdPyDH5vV45Y0GgAkwn2FZgQIQhJ0XHk657BKx7yVYYBx26dhNAmMQPsxSEgnBi+iyVvEXRKJNhht7ni3gw98dnXLyMAe3TijDW28rdI0jbIDH2/5EDyZkrx9OmWVC8pUk6SKZjcwm015sbnh4y/X1ENApynL9QHba6rDgdlsTp1HNeZssaSqGtIiZb/e8vL5La+vdyBgvpji7EBZZngt2LS3nK4WPLvecbltmZUpi0kOWrPvLOu65WyeUWQpl/uO26qj2npSLbjvBJ4Ov0i5uMiYzufgWpq+J7MFeapJsoJquMHgEcExWy0RSYV1W5q6YX5Ssjqfsbm+pJwsmU4KWhcwWP75b7zJ05uPqcYNq4yDLIxWuF7eyXa1Ugx9T9bc8E/feJtVIfHMebV3fHJdUxuNHaDp4VXVU+17BtexubkizYrxhk/uSs48L+K4e8xHtc4SuiYGACRJPE9yLG2Gnq5raeom6puaKso03HBXdfzqh34U5dsQjR2di9tYqRVGpHd/ea0NalQ5Hg/4kVYXfLzGlYwwYO9i46iVHnvsMDKeutgbENDEyUFwbnQpiZi2i0CO2SexYTakSTISKzxXdcNnt2swyWgW6caFU5RDiFGWut/v47bYg/Keb72x4NuPFpxMJPNJwmA9Vzcx4mJWKN66N+PpTcOrXUeRKU5OTlmkCZc3W3SqCM4ixZ6urVmdL9gfWj57fosPhjyFQ9WSpnq0JjrSROOI+wGpPFMdHVP1EBmzeWa4kIqmd8xyw01leLVtOXSWl4cB76G3AcktzaHlYpUDnsbeYm4VRbomNynWDWSlwQlPlidMp1OauqepG/ZrkMIzKadkSaAsUraHmnfulXznzRP+7NM1jPJdazucjXLfMMoTpJI8WOT8x79xnzdmkbI+WM+by5TzQlLVHcIbkmxCJw0/fznnz5/ccLmt2W1vx4lQ/EyEFORZEXMo04xiMkVqiR8U7eh3PnIP6rqi71qqw562PmCHAQhjTHwxJtl9DZVla6N1zTofFytdj7eWgEAqgzIGpaL2IUWgGBlC4+2utY5soXG5IUKMxFZCRiuaHm9xrcYbPGpfgvMI/TdqN6WRiYnLMBEfAiEgMZFQHTGP8OnVDbV1MW9GCgY73BFSnHPRixkCw0g/RMBpkfAPP3zAqkxJjWeSJ9R1y6zQ2MZysSxYWE9jAy82DZ9c1nTDJe+dT1jmmi+eXnN2MeVsPqFYTthsav7q42cgDLebivk0IwTBzfWGNDMUk5KLVcGXL9Y8Xmb8/nfewEnBn/z0S673LbNJhhKKaZ4QhKBuOt44m1ANni+vD7xaV+y6+PbKtj1N7heSBwAAIABJREFU17M5tJwuCpRyVPua9956wNXlNUYrJl3CrOujaT9oDvXAdlshlSRLBdvdFoRkkmkaJ9nsWuoRPRrzguKeJkqzj3onS5Yo/qPvv8e7F3OGg0VIjfWSpu4ZggSvOaw3mLRnspzx3UcL7i9L/sc/+4TP9wNusH+jMoCD7amqGOOUpAl5OaEoIzA5eB+Xmt7TdhH6MHQtjEyEWNfHuEWpjjDpX/HQd70lBEHTdVg74EermCSOE7UxaBNtfAkCMZYt+HFzh6C3DiUEUVTsMTLqeQhxUhOkJNEGXcbpjnMObx1exkZMjYYB71xslkelpxTclT/BBzoh+Pz6NkphRXwbDH10Cdkh3rBJltGPljQx/t4Hi5yTMsHagcXZCWkSG+yT0wnZEPNyto3Fhx1FGl+9X9zWIOCd0ykidLz70RusdzsUkqvLPQjF2emc3jrun8949+1zejuP+e5ENuubDxZs9zV9Z8HDP/z2I37y+SXN4KmbgcwkCBFIheJsOUcowbv3Z1xtOz59seHl7YFXu46mV2yaA7eHgTcfLDFZzrNXG+ZlSmIkr68O7LeS1emMpulpBotKUvaNZXay5Pnra2wQvPHWI4qQ8PrLV9zuu1hOEDcyQohoflGRami7gbJI+fbj+7hDTy9EfNubDBOIYavdQJAp280GqSXZRLAqEn7vg3u8vt3SCo3RCVKOTS+CgIjZ+SNVsGmbOM7G44fY/8X6/0goFCN42dzJ4JXRqP+f6Ju/eyPrGf2KPYONY6BE6WjiTlMSk8TUWBEXSEYJjJD0FoyRcbHRD2ilkSMpzjnwwmOHEc4mJKAJIR7c6LrvxwjtMQeEgJYKqWMdL4W8WzvH4H7F633F1b5CJQYhYyqWtQNhNDF7QGvNfrcn9hwC5yxlarB9T9N6nj1/SRrpMqSLOVpaTmYFXrScTHPyDPbdQN0JXh96gt/xxmnOLz59ge07go/5K6enS7q+4dsfnXExz2mqLeV8QlkUbLc71jd7goBimsdcyN6Tes8//q13+OzVhr/+5QsYumitlAEYSFXCMjNMs4TSCN6+mHK5aXn2ao33BolFX1fkRvBgmSNFz2p+ws31nmcvDyy3DbNZiUkUeZYx9D2XVztWqxWHpufl6z2vKs/Pnm94tt6DyqMPdYxH7PuBNIm3qnWe1ChSqaj3DUYJtNZgJDKLg4Vd3eGFQqiM28sb5kJifOD+NOPbb57x+dYRVAJCjVp4iZBqjDb3UU9DwNuWYZz23E36RFT+BilGmXIsnY9ANvG1Dv0RZzIeMKMNQupxoxape1roiMQMdpzrC1IRb1JrLWpS3NElpIhunCRJEGIYnTnyrtnVY0OaJMmds57x3yV5jgiRNetG4VEYpcpCKz5+9oKOEJ98rbD7HutjvWjHtTjjnwkRb9zBd1zvW3oACft6wGUJwTosdfy7SDhbZHw/v8flduDZzYGrbUXTWxoPnRNcbiq6ruHi7AIjJXXbcrFK+bWPHnFzucF2lkQr/NCTp5rV6Yy67rC9QynBwwcLLq/3rG/33F9myHfOePpqhxCK2/WG7TCQni9JVUISFI+WM3yAR6uB83nCL77Y8PlNzatdze9+eM6vvbXkfFHQ1j2//v4FTbjH1aZhfWjIvSCVgjTPmE0z9k3L1bbh6W3LZQ2/eLnn0Dqy1JIk497Fx2waiBM4bx11O/Di+pZ5COSpJ0sFUga0AzPNKPOMbdOQlFMOtzW79ZY8eIxJeediwQaPVemYhBBPqRz3A9ZFOXpbV7RNfRfKBdxNBOP2f3xTSDVKJiLY+2ulFv/JTz7m2+++EZ9kpVFKY0wS0wvGtKk70uf4h41YdxM3p4g7Y/jx0B0nLUpqhHB3H6o4jjjH/9ZaG5tPGflV3lsSFePeQI1LLEnwgU3X8+R2i/VQJEnkV/X1OBkQd4xS56JP/lj+eCd4fWj5+PUBQcm9iaYbLG3TsdlXrE5K0iJlUeZMQ2BeOAojWOSGq10TzSjOMssKEpOyO7TUhz2PlwXvPHpElqfce/OCtunwnaPrY19kEoNrLbttxen5HBU808zw6fUtZ+en3FstWW9b/upnL3j/vQcoCUWRkiYptrP0A0zKiD96uMz51ltnfPb8lrNZxj/6jQeUqWB9tWV3c2AIljfffUgxUcjXgab1HOpoqXu9q9geDqh8wlXteXLT87OnG2KpERjrmzsbYJrEFtEHS9X1/PjJaxZSMc88kyyQpwE5OBI8YnSDBaXJZyfsb1+j0ziPX5UJ08zS6CISVY4GkRBwQ0/f97RNQ11Vsawez4ccHVlHtqz3AW1G8oiQUQTZ99ivg9/5b/+3P+ajn13w+7/1LT5863Gk2umYK6LHGj6I2BDGpjNgBxdFXt7f1YMxti0ad33wdxmWR+NFZInGaL3ExBm+9zEpTIo4ExbxucKPERA+xNz2oCWvriuuqxqdZkgh6PqOwQ5IJWM4krWU5YR6JA86H+itxQt4ve/4tx+/wtszTt47Yxg6qrpjUqaEEN86iYokcd/UTJQjO0lYFIHLdUvbW7b7lr4b2O13fO/9h7x/XmBkrFOTRBK8Yl+3SCXp+o7ZskDpKTqRtHXD+haev7jk8aMVN5u4bX3r0SmHqmZSJKRpdC9JAUIr2rZCyQwkpHlKaR0n756xmuXU+wolEpanc4oi4/Wr1yTCcZYFFm+f8PMnV7y67qidBKnZNYGubfnlVcOf//KKfiQBHlPH7r67o29JCELwdM7zy9dbPjxb4FqPnwi8kySpo/MWmSdILem6FqUMQmm6uiEtclJtWKSCoAyD8wgv7hZWu+2auqnou4EwUmKklBit77L9pYr6LMaSWSmND5YwZl06/7WmN56ffPaCT758zfd/7X3+4He/x6OzFX487FKAcDGRCm+RIYYLWb6C/h6lCsfNWRiDn44cqVjGREeMFEdDWDzciTbxVRXCKDzzd28MQnQ7IRRPrm+wIZCpIwnF4kez+FGRp7Ri6COCvXeOwcaJT+/g2brhR09u+eDeCbl37OoIcNPCg5SIXTMa1SUyBJJEs5CGzAiuNg2XNxVCwtsXc37zG2cUaUAqQdd05GmJNobJNEUKSdflHHY1RZlQzhJOL2a8fH7LZtOS5Rlvv3lG0zmCh/OTKZv1AVGWyAK8VOzqmtlsQjHJGPphtFkOKAlSB2bllNRIqkOFEoJJaqhut8zOlgTn+OZ7D1mdNvy7v/iCZzdbiknOYAyfvb5kXfXE8cBXN6mSAqOjHNiPkLNeCJy1PNtU/K8/+iX/+L236DpPM/Hk2ehf7jRMEroQJQ/OO2zbU1jL4D2uGTgEFSUKIaYt3N7e0LfNHTwtMbF/ZJSziwBC6zviiZACwujYGqIsxbuvSRcUIt7Mh8Hz73781/z4k8/5ve98k9//rW/x6HSB6G0ED/iA9GF8KsWdpgbi+DH4GBl9xKKEEHXQgYC3URPjR0fWcVNqlEYriZYjUVrqu7cGYzQIEmrheLq+HWHDUZXZjU1PGEOglPzK3Ox9JBQ67+JoS0DvAr+8rvjv/uxTHk4UD+c5DyaS5SDwmz1FWaCUoW07brd7ylkJMpBnKZNioO0k3/vOm3xwb4GtDkzmc3SWIIPDNh1CjZOstouZm7uKerujnM/ph3h7vf32A1zfoYRgeTJFbBs8a55fbngVdty7OKF3LWcXK7SK2TpH9KS3DiS0bc/ODayWE9q2w2hJWhZMT0rQklLl1G3P+arkn/3OO/wP/+en/O8/veTlYeDVtsOHry6dEOQ4kxfE5OA4JpUyfs4+BIJS/PjZFaVW/Naj+/TWkScpRZai8wQxZIQkwedQtw1SQF21HDrH65uGG1/d6bi67m9rsZQkylFUTONQOjarybiX8eOZsz72dc57govpXvLr1PRKBkIQd5fqumr4n//0L/g/fvRTfvDR+/yT737IG2dLEik4EgalUHc2Pu8jF9WPUWOCCDbzx42rVHgR5b1RwnqUtcZbPT4cMdXKigGZZvELEALwuAAv9zVXhwaTJuPXJej6Nj5QPk53TJLcrbpDiBmZcaJz3BQHDtby5a6h6gyXe0t1UXJbdbx7MUVlkv3tlqbr6YIj8wGtFcZEdefb9+a8fT5BiIHTR6eEENivd/RtT7komS6n0dGkYsy5NprdvqZpLfYwIIiSB6EVg/W4ukOJwHtvnXFxvuD5sxukEPTWkGhwrqeuo2HjSP1ItOZkljCbZwQs03lJmkbjTTGJW+gkSbjdHZDek5cZP/juW/xPf/GKF+s2bsuJF1cQxNKRrxKAj29NbeIFYr0DBMEY/uSTp0zSlA+kGFOmB0xvkG0Lacoiy6JllEDd9Dy5PvB03dPLBDnqeY63c/ThKoxOSJKEoI7llRyzdWLTGogXWHDhDsyAGFW7+mtsZLVJYtqAdeNfPuZQbpuOf/Mf/oo//unHfO+Dd/iDv//rfOuNB+TJV/87O0S+bHDhjiBunRud6jGqTQiJEEeR2qidH8NdQ/B4KfDiK+0HjNknPqoyg5R88eqK3sf8SalUbHzcsU+ITZBJzN3mLhDGGLw439fAal5C8JSJYp4ITmcpn93s+BLPuhlYvLjmJE9ZzDJm8xInIzW87VpCcCQ6JzhHOimQJqGpOvZVj+0GkswhpKZpuyh59QEXQJmUph04HBq898xnOZOyxKMYRlxp5gMtLd/+6BH7fUU30sh1mrJe70mTnKZp2B8skpYP3z4hSxVlkRGCohtaEm0iHSRJCVJyerqit9Eb8eCk4Lc/us9nN5/RHb/e8YAjBB5GEglAwAeL96Obre8QQJbnXO33/NuffQbA4/mcIhnIeo1sNOnEU3Qdbd+jleR6u+WzdUdHHDj4MS/nGPFhTHL38x0BEROeOFa1PjrWo/suxEtOjvC1o1T9a3FkAaTSmFGnHjNH4gF0wbPvHX/0k4/5048/4e9/9A7/6T/4bT564x7yOFLSGotFekWwY/Sdj4IyhyfYr26SWH6MGnfvxkWIxdo+zoCBfuiRQo4jT0nvA882W4JSCG0IY5iTtXHEdTSNJ4mh69pYHglBliRYHxBu4AcfnPGf/d67nE5SilQzyTVKwXrX0nQDXd2QGcmD0xn5NOF2fSAxhurQRK2QliwyBcEilaatKpTQ5ElCaxnTumKZ56xjNptTNQP9bovWijTVzBZTlAx8+otntPXAw7fu43zAJIrT1RzvI2hNetAuhmKlacp+12BdYHto0VpwqHuWs2gX1FpifSDNY4OZZiloHdm3g8OElMHt+S//5a9RlP8faW/yJFt23/d9zjl3zrnmN/WMbgAECRAgwAECKZqyLYlSWJQjaNle2d777/DeKzvCDmvjCIVDCtmMEGXRpBmiAYu0KYAU2I1Gd+PN79WY853P4MW5N6ve6waa7k4s0K8qs6oy7+/+zm/4Dgn/4796j6orhXd68Q5ch3UJlNx9z3+2jlZrb5iXJKzrhr98eIq+bTkcD8nCkDQNaSSYs1OqusTJkLON5EpHGCF3GJme+yw7RCVCdJnb93M7/cpdBcFuU+t7ROH3Nx2S99Men4qyxPb6iTdIHV0DCj5I87rmD7//V/zpux/ym1/7Cv/wO9/kzmyAzxX+jclOktn6d+MxHUbTuz73WzmvUdiD1HwJ1P++Xqum/wDmRcWjiyu09qA4h6M11wHfCw8J/FZWSkljLHVjUcCrhwP+0994my/fHRAp5xu/YYyQcLKfYY1Btw3atERJJy5lIqIwIgoEZanBOGR3MjlnGM6mPP/Jc6QGZ1qMcywXG4aTIdFwgGktuvEQ7f3DGePpoKuZHbfuHu/m0FXdkMQhRVGRZKmHeqiIpjXUraasSpq2ZbUuOF1sqY3m1776Kk1ZY23DZDYiTUJUIHF4XrISFqMbdFUjg5DJICNzjv/8b76OA/6nP/gRlbmGdO90Ra3PyFEQetQj3iShaVo/3YoitHMYIThdrWisYZImDHRC0mpWdYUNI+aELESAURFS4kGKQuxqcuvA1N7SVAhfwuzAiu7FWj0MvXpaX32owI/TPUT5czCndgui7iHpiQXduLGrw3pa36ap+d/+zf/L9z98wH/1936Lb7x1B6m9TonwBaKHHnCNvuzn8/2R5r9Gp2Ard81xz8zytW2NMYYHF3O2bUMUD/yGDtstNvzPMVoThb60cdZPh/x7cAwjwX/wjVd46yRlf5Z41d+uV2jymihJvOJBqAhiRRSHLK423QcvsBascZi2RaeC/cMJMojRVjAYjVCACHwn4zVjHGGoaCu/Xj+5c8hkMqSuNdttBQ6SNGI4GlLkJdYZlITZZEjZS6i0Tbew86O5Vnt0YmNXCCFZrVcc7+2TpAlJEmNMS1u3fpjgWk/sKXI255doHOODfRwhw0jxn/3mF/nwyZI//OHznQzKruzsELbKepMLaw1SKGw33gzDEKxlOEhJYkWpa2zZUumGgRsSxWPYe53lxvMPwg69a7QBJVBRSKJ8pt5xLPzyZ0cb7fH5rpsoxWFIECWdgGtInCQ7bdTPhb25GfDWWj8T7e62/g/ZZYSu2XUCHlxc8N/8z/+Mf/Dtb/Lb3/p5BhKw1jOfcH70dANv3y+OlJQ45w27pBQoFXXNZz8Z8kA112nYP7pYYHvxThy21Z21/fXf3td5PQsIHLNU8h/98lt8880ZgzjwQzopCZQHPMnY44NUpDDaUaxzGA7YrnJvDhZb6qrXd4nYn028z2s2oMo9gUE7y97dO5w+e04YSqpSUZYFaZpxcu8E8DdMGCrCUFDXHgXa1H79nqYJm82GJE19dm000vn34aSkLFpW25qz5ZYkknzn628yyCTboiWMAoqiJlQSUxvqvKANApq6IcoirHNsNxW33hhycbnm/LJExRl/6xfv8m/eP6d0vXBvx4buPk/XDQZcdy208fuUMAy97qQUTGeJXyTGQ+58+ZcYnryByCZUTcv28cNu2+6xVL5krv0p3jaeV6GU1ztFIIWHJdCdpCjl9fhFh8FSAXGa+CbcOYqypG3rrpr4jEH/QsB3jawx103H7g7sPhzb+c4KKcibln/yf36Ps6srfvc732SWhNCabsnkydxCXteJfZD62bDs4MO+vKFvdrvFiHWOyjqerbYEYUAQxjjh7X96eT5j3O71vdoaXQl0OM44HISYtiOSGI1wDhUktNp2hsl4CW4HujFU24o0jojjiKYxmNayLXKOD0a02qAbgxxAWzdcXV2x3W45MY4kTTBWe8fFNCVMUg/PFgIlFFI4sswPDJrWa20Kgs51cLATqRVIyrwiCEOUgm1R8d3vP+ThZUVlLF9864RJHBOo1JPgqxYRBdCJSi0uN5Rlze3Xj3BByGo9RwpFHCqySFK1Dd/4wj6/8NqMP/vJEmc9dS/ofj+AxtI6j5lygg4m3s3ulaTBkcQxcTLl1W/8Jodf+hXq1lLVBdViwf7BEUEQUtc1rjUYfIKzWNq6wbQe1GgtnamaRmiJcXg/3w5kGEhFJWs2G78/McI3tz29NfiUuv5T5/TO+QlhR+nwBHFnO0aV2GV3J9gdhf6JDoPjj3/wV7R1ze/++jeZdSrFtpPoC5SHE/T/7mWZVRB0MOVeLq/rI4zf8lZtzbO85iovCSLPHPJqtlX3ofmM1LOu8jzfXThrLOu8YpXnvH1r2EGfA+JOXbeuNGEQUmzzXfkWR547oJQiVAEyUqSRIYrGKKl4crZgEIfsV5oHD0+5uFryyp0TyloTBZKn95+wf3RAdDfxsil1iejIOarrUawzpFmCs9A2vnGrKs9Okx3TbDxNvU7o5YJ3XjmkagXV95/w/tM586Km0UFngSQxxpIXZRe0kuXCs4yshtFowBtv36PYrBHGcDBOWJeGeZ7zna/c4QcPln77KSSBUphuXHltuuA3oqbVKBng8EKv2kI6nnH3577F0TvfpNaeCohIaZpz9g6OsdaSak1TV57/3Gmc6tSL/7ZtTau966Pt7FmFcwQd4QfrqYpKBXiehSfICymI4+jzN7I9+KIPbvA1lRAeAiqQu5HTLuM7T/XrX6Cd41+/9wFIye9866vM0hjV4XT6Uqlp2x3IzN9sYte0ONx1Yyv92BOheHK1Qjsv7BPgl1/eTr3f8HpzgL5GdH0mEI637+0zGwSME+VLLi/H6Q3ChMRoz/jXWpOmaQdSwxs7WIe1rQfhCcGmbFmsKuai5tnllufnC5I04dnVBqEi5ucXbFY58WBAvFigW2/UMBh6Br9uW6y5rlO3m5y6rAmjEOcMm/ncE6gDRVU2SKkYT1KCwPFrv3iHL799xL/47gds12vit/Y6cn2AQFPlOTWAFWSDkP39g66JlQyjiLbULFdbrFPE6YBRqnj9KOPVwwH3597L9+asfofBieNdWVvVFUkaI4Sgai3B3h1ufembWBkiXIvWmm2+Zbq31wnB+uQVJ5kPemMw1vvC+obWe8Jaa3YTQ9v1Y65TUb4ZHzJQu4WVX9yqXYL+TEHvXB/4PtvvuJIdaN/ecHy4DljfTe9mpcILov7JX72Ps47f+ZVfZJwEBEKgrUFIv2yy1uzufGAnyC+l9FzdflCLoNCWh1drrBDEceqXZ9bvABz+xnB4KHFd17vyzBrDNI04GUdESrFeF8xGY+qyYXG+omkajm8dEAZeSaFt/OZzV98av7Noas16W2CQFFWNsQIVeGhsmvmpT9M0nJ5doJuKwXRC3Xo4dVU0XJxdcXxrQnYr9sezNgRSYG1NmkWepCM9pvzyck0YeyzN44ePGE1GHBwfe4dEo5kk8Pe//Tp//sNHzOcbDifZjgbZNJosichGMdNJimk1Za5pSsvl1YrJ3oQsjVFRTK0F0jpeP8r4+7/8Gv/9v/oRtXE7sFd/jXtteNs5fHv4iMRJaJyliYZY6ZWlhZbkeQ4OBoOBv07OYZX1Aw7rkMoSuAgVhDsZbq+fartFpuliyMeGsQbhfLaX3QY+kL0LeZcwxecI+uvgv9kY+MmF+4Tv90ybvq7vN68In/G/+94HlLXmt3/1q5yMUqTzujgS6TO4FLTWIEy3fQsUjfHqCVg/l610y6KsOV9tabQh65rhtq7Qpr3xV0Iv4N9fOOcc00HI0TjBtC11EyKFYLPeUG5ar8lZtQSZotVemUF2Jsxt46Wry7IGKUnimFVesloXaG25dfsEFSjGwyFtU5PFCkxDU9PVuQGHJwdko5RslHH65BlSBuztT/AdnO+Z6rpmvc45PNijVYr9kz3COPQ2OtYxmUyIk9DTIDsZvWEsee1kSlE0WOto6ppBFjGajIkUlGWFaVrGo8wjJMuaOEnItyWTULHONzQypHISZ+DbXzzgL+4f8Cc/usKv9a/Hxv2pKYQn4uvWT5aU8CDASkNZlGhtSJIUnCNNkg42IrpyucfMmF2sJEGKDe2OIOJPbB9njdU7rNUOCCe6uwO8irUQeJO9n72N/WsH/c3Ads55QIHrVX5vljb2hcxw83VCQGMtf/rRAy42a/7hr32Dt24dgvP+U8je+UKhnUNYRxh0K2/tEZitNdS65WxTUhtDFEaEYURrOpt0bXYYGyFfFPFHCNq65Wg84tWDAaMYYumo6wqc8PaMYch6kSNRHcPKMhim/hivGoz2F72pG5RQvmHUfqZeV40nfMQxTWuQwjGbZIxnY+rSO63r1rC6WvPw0SnOel3OpmnJhqnHANUNm/XK6zPqBq0blBJMZ2PmF3OcFDx89JQvjd7whnKVwThLEAlGo5iLqw3n52sm44gkVujWoLVFoYhjCGPFdH/MtpyD8L0FDuI4ZrkpkanvcUyr2UsD71LeGsLoemJ3HfTihaTSv5fzi0tUHFE2DUVVEQQBSRJ7rUvbYWOE95sKhMSYXeGMpBuQyC5tuW5J5bpBh7GeWxy43VDD6y35CRzSj3X7CdpnDvp+nn4dvB0ykhe/3ge2MXZ3zPTZtfsuQnji94PzOf/4D77Ld778Nt/+0uscjQd+rt3dK315YruJkHO+eW6tYdu0PJ0vcUKhlKRpGwSuyzjXTB9gJ/LknJcoTEP4pTePOByHmKZmlIS0eeG7fWd5eP8p48mU4Wi0a9aqsu3eh8d9CDy54vJyxXa1JYlCtHZs8gIFrNYVcRyBkTjdsrc3omkdS1MjgxwEzJcV2SBlUzREUUBdeNdAJb3i13AYUZUF1npYtG4btNbsHe6x3a7JhilN1QDWEz2UBBXy8GwDwvtQJVGENQ1a1xwfTAjCgPl8gzWSi8sSJRVR4KdXIvCEeyklVVlQa8HRbITgKUb7EaFyEHXSfm1Ze+cQpTDWf95KKnANp8+ekW+3pNmA5XpJEIaEUbwT8rLdZEQKiVDX2d4aj4WSSiCd8lm9y+7K+XGzCxxSC++bQD/29lvbtmkoNjlFvmG73nz2oN/V5c5PZqw1L5Q1L2f0m1CCT8z2XanjBFyWJb/35/+O95+e83e/9RXePj4g6rBkTviRlWklSgqccDTWUbWaVdlwutzSGkOkUs/kMa2n63UQBmMMcRx3Kmx+MdQaw/Eg4mAQkkUSGSUMwgDXOqqy9AwdCx989ITxZMhwGDIcDb330W5MC2EYU+uWIAw4OjlkW2o22yuulrmfG0eBtwAdJIQqZL2sKeqGuqkotKauGi9am0Vs8hbdbtibJKRZTJFXrNclSlYIyQ4+e3a6wBrD7TuHBOEtPvjRQ548PueLX3qV6XBI1daE0hGnGe8+WvDVNw9IVwXDLNjZ2lgE8403ggiTiHxTkmVeZfr5+YIHz5e89vYbTKYjVtuK/UlCKLv6WvvFXdyZYzhjCJQiCiPquu6mTf40yDcrnj66z/HtV2itptgUBIEiTTPa7tT1RB6xo/1Z200J6TFRfQNN19/cUE1WCino6ISeCVeVOVVRUBUFutU3To/PGPQ9lNdZr19JB+HlhSx+/VAd6OvFE6C7ATrQP86/ae0M7z5/xsPfv+Krr93hl995g7v7Y4aJ8rRAY2i0R0TWumVTV9y+DikwAAAgAElEQVS/WLIoKhye/e6EQDdeCflmOderqfUElVa3bEvvQFKXClPVFM5wfLxP4gRyT7LczGm15dHj53zhC/f8UW49/sh1YLlslHHxZEGSZYgwAldy52Sf9x6ec3659NJy1lBWhrdOJhxPItI0wiKZL3MCJZmN093pcnW5ZX2VE8UBl4s5d+/dRghJmsWMJgMuzhacP7vwGTBKaJqaJMuYziYUlWYifMPuXE2WhfzwJ1e8/coeTgUsNw2R8qer69CLCMV0b8h0lnn+g3PcvX1AECcgDEEUMR7EvHYccTBOeLJovMCtUt52U3tnRn953Q41u8NHVRVRGLHdbGl1y/5sj/ViQVNW3YLR0/uc8JMX53x14LoTQ7ctxnhCvzEW5/xw3iFom8aPVaxB1xW6E4VtdEOjNW3d+Bvm8+je7CYyPTFwx9+4IeD0UuBLeY2GvDkz7bE1u9cLR68vkbcN3/vxff7t/Sfc29vjq6/f5q2TGdPMKwJ4zIzmbF3wZLHF0CFApR9Pta2v57lRw1+vtO1O1i8OA7ZlxaMna9569TZPnz4lShPGWYyMUkajlJPjGSe3Zsz2PVHjyYNLNmtvE7/ZLLhaLWkbUKnG1pa6rJgMM4aRQ8zGPJ0XLIoG3bQ463hy1vKFexNevXOMc5bxMGF+cYk6mREG3gDu9HzJj59c8uUvvU6YDMmLktOHZ9y+tUdVWQxeO3KzrhHSk1iiLCWbjMjLitY4WiuRQnK2qnn30YLxICCUFpUEhHHEowfPKbYVUTrk0ekCsLx2e59VXpEm3mk8TiMq41UkAm1462TMk6tznPRLPa2bbop3zWjqsVP9CLFpGh4/esg7P/cLLK4uyeLEjzOrkiTx5g6mUzYwncRMVTfk+YYoinZG1h5NaTFaY9qaoqxo6hrn/4BuquPRl7408jcPzu3i9DMHvezecIfi3E1kXg76m/V7n2X7D6Pf4PYjq/65/YSlu7OodMsHZ+d8dHFOFgQcDgcc743ZmwwJAsWmbViUfnERKT/KlNL5rZ6zBN0kp29+vQaPRVuP3Ll3ssd4NODu3h5CSvYPjlisNpwcTJntZbz26gFhmlKXJVmi0GXO0UHEZOK9Vo+PM+q6pmkseanJa4NxjvUmJw4kw2HMKBMMRvt86ZUD8k3O0/M1rx9nHB9mjGdDZKBYHwyYX20IwtBva6OY2gguFmvUY8mtk0MOjo746P5T1puWs/MloZLcvbvHcJxQtxoVePvNwShFt5a6tYRxgpOKp1cl801LolqUjKm1JYgjTg5mrFYVurVcrUv0kxVprMhqga4dm3pD0dSkgyHWtHz1tUO+++4ZusPZ38S8O6wn7bgXUZDOWpbLBcvVkiLP2azXjCcTmqbl7OzMO0t2DbExZtcIu47U0wMHvQdZQ9vWPti7nYFQPmRbbZiMJkRhiJOeH6u6uNLt55D18+pWehe0/dec8+jJm7W77TrzPgOITyh/Xq7zbz7nWo/e/68wLQ+WCx4uF/TkkCyJuHPraAdDddabEesdPcwfkS+rKRjnmCYRJ6OIqmgoB4LNekkYxRwczNibJgxiR5CEPD1dsrhccnw0IEkT0uEQUVa0TQ3OS184GZMWGjEvgIh0lHF0PGaSRgwTxWw2wlpHsRnwjS+9gnWtL08CiROWO3cmKOkIIk8XtD85Z+9yxOWy4tHpQ26frhHOUOQlwklaKxAKlkXNpmkYD1O00QwHGbEJKMqSy2XBpqwYDgJkoLhcVYxjQxIFPHq+4vL8kp//2ozYSjaXW9rWcLbYsD8d4fAKEU+eL4iSmBZvBO0h2q7bvbgXrpnrPt8+6K97OsPpk4e8/aWvECpJWeRs1itWqxXL1cqPgpXykxnjSUM9x1p1mbtf83dl/863wAFWdKYR1hLGEVES7/A6/e/ve4fPFPR9cF4Hez+Pt/SQTl7Yfl2PkV7mwV7DQm9sW28E/CfdGP2N0M9tkzQhjmPqut5ZfZqm7S5OXwO63SYWPJrPWhglEaGCotU4MSROMoaDhEEac3q6IItaDg6nuGJDGjiSJMEYw2qxJI4TqqqmR4Ke3DkiKxukEJizOQd7GePxkGZ9RUiLayuqsiIIvBdVnhdUTcOdV06wxrApN75D1C1OwttvHTOdDvnj775HFIWsi4o4lBwc7aGbliiJEALyTcm2rInikEEW44Tg6nLtF19WkMYR+9OM03nBclNjWkscSeqm4tXX7rDNW1ZrXw6lacrV1YrFqiCJI1rdUtWeuBOlkGYJr98O2BtFnOdmt40NgwCsxdjrMnI3xpQCDKwWV5w9e4xSIWazpmkqirJkvpijlCIZjAijBNFpJvXNmOwGHZ591ze7sCuL8ZtrZ51nR3XalsPBkCSJoTvhr5PgZwz6PshFdwtKSddt37z72QV/v7nt19Q3b5SXT4Bd1ngp699cfe1IJgiSKPZqCyr0zVD3QXju7cczjrEeu0H3tcW6IFUp58ucN04G7E2H6LqCICAYD9lUlotFjhSKUV6RDmKMEJR1xXg6QQhP9l5cLqmKktFwiDjaw+qWZrulnG94+vgZ46MZMobhbIaSSafj6ceeRoOQitEwZLMuvH1nUROqgHfefoUHTxaUZYG2gueXV7zz+l2iNGSVlywuNMtFycHRDKkUeVHRNI5WGwbjCU/nc95/tmJbam5PBgyjgG1hWTSFt98cK1rtiMIEQUuaJWy2OYMy5dbRCKkU69yXG2VVESnHL3/xDr/3Zw+wXfIxPURbql2C6xdKSiq0MDRNzeLqguFkRtsaojBA296IO2KQDgjTtIujF+VfJIDzDjbXRh/Xye9mxdH/d1kWft+Tpd4I4hMGLH/toO8D+uURZF+z9wHqAXkf34bdDOrrr13/TTeXHNfP62+m66wvOlx1FHoltCiKO03M7nV4AsfNG6tpW7Q1tMaijaNs/BG+rVrOl1t+9RfuItuSfLshnIy4uFh6dlIUEypFUdZIBdlgtLswTdVQlBUSwfJySbVZe5xOmpFftjz8yXPS4ZA2t9w7PGB2uE+5rYmzjESntI3h4tybLqiuiW2Mdy+v2paL+dbX+M5zAcqi5dHzS6JYoUIvYd26go8ez5mOY5Rw7E/HVK3m8emKP/iLM/7yyYa9QcyzZc5+NiJzisZInjxb80Y0YLsuqepNl7wsQRRz//ElTasRVjMaj0iylG3dMEpDfv6VKX/w547qxkaW3fW/5lX0RCCDP1lPnz/jbhThRIh1fg8QhCFhGDMeT0hHw+ubRXn8zE1813X/yAuB/vJU0DfGLVWZezvRKOrgyJ8j6K8DUb4Q1C9yEfvn0AXrdba/udyyHc/rp62Ke7HQHrvjXE86gTQbMBwOvP+UUkjh8RZOBbu7SHTjO6UU2liMERjrsddFY2iFYDIM+Ll7U57ef+5NGIYxKo0ZDmK26zVFkRMEkiDM0M7rpDtnuTq9oip8xtfastnUFAVMphlSKoJUcO+L99g/2gehKPOS+x+dESiJtS3T2ZhNUXJxvuL2nWOM8dltm+c4KUmSkCgErQ3aeB4vQnJxtWYyGRBFgqpuCMKAJ+db5tuKURaSDYdsW8m//MFzvvfhFQZBUBoKLVjmLcMsJE0HnM/nRPGcYRZzNB6QZiFl1XJ2taHUlrPLLfduj5lMM9I0ZmAi8rzlYJIyGyU831w7z/ik5InsSOG1RzuWVR+5q9WcwdWIMBkQhX5qE4cpSZyQxAlZltGfzghJFHskKz1kS/aGadc7nr5x7isM6TEvHnNlDM5oTNtg6urzBX3/V7x8h/UBfU0bvBnQLwb2yyjMfmPbv170DQzXX79JDRMITm6d4KSgNdo3PJ1RbhB6FzxB628QgZ/F1xpjZSc6JViWDWfznL/x9l3ybcmP7s958/VDZid7WJwf04Uhw+mEOI5RoeRquSEuGsJAUVYVaeYD3EjHnddPCKKQNE14/viC4XhIIiHPC4Ig4f33H3F0OGZvb4JSEdl4SFEtadqWvKqI4hiLQBuIoxCE4/W7+zy/WFNr/1lKJNPpEG0126L0XAYlsUIShglFo/ng8YLLjeHHz1c4JFZAri3zbUM+jThf5Nw7nqCCiKdnS/anA2bGgPCjxKO9EVpblivDs2drptMRQdBQNw1ZmuE2uqu5O1gAHtxXNzVpEHXeBXpnrNxjvYqi4Aff/wsQ3oJTScloPOHrv/QtkIIo9jpAviT1qnh0QMZe4gNkJ/0hdo7x4E2Te36tE35fI6TfpOm25nOZMvR4+puBe7OxfYFE8gJU4eONaX9z+Dn+tQPcbiza4/a7G+DmzwvCgKPjW6xXVx1hWOxkPZTyUm9t45dmVhvquqZuW7Szfh/gHI22NJWmWNRcbFa8/eYtDo5GVNqSBgnaQZSkKKV2fljZeMpmvkBKQWsg0I7AWDabgkBJ3zw5xfnVhm0NWRrxk/tPuFpW3Lk949W37rFcb9iUNWKTo63m7hvHiFCSJCnLVcm2aCkqx2SaUVQVxgnWpZ/xp6HCCC+6ZTSUjePB2YZGO0oNjTYU1QqhQl4/GBLFNU+XlT+JKs2q0igHVdNycvuIq8u5DyYVsClaim3Owf6Uw9kQpSTbbcPVPAdjWa+3HN0a+qlQ0aJkuCs3JKJb2ilEqDCNxfanrPOJRnbIx6KsO69ehZ7PefL4Ebfv3fNlYRYTRhFN48n/Qiq08QOSQPWKxAqkvKFR6bBBiKQvgfz17VGWXvnsc9b0feC+vJD6pG3szdft6vGXMr4/EXrRTre7Cfz3ZPfv/jX+d0zGM4aDEavlJc56ESApJVXtPVzj4bBjHAHOee3IjnwguomSFPDmnQMMMJ5OGE1Saq05fTpnU9UczEbeVdxob7mOw9UVVknGezNvy6kNKokJ24bVcsuTsxUOT36p1gUf3n/G1SKn0o633xnQaG9VZCycXS5IIsX+0QwhlWdfOcemqqmrnMvlhsYItAh5eOm5uEfjIeunW87mOdpY0jik0Y408iSbaRYzHSa02nA0ldw5SHi+inn36YZV1TIvNFmg2JaagwNBHIfUjebZ6YLZbIhFUNUtRwdTwiji4uoZH9y/5OBXfw69qVhsStIkZBAHGBNQN01XSvpTPlDeNrMoip3sSp+lHd5ATdYt0MGCjeW9d99junfMz3/ja7TaEAahl/AQChWEu5Oi3/IGQUCvoXRdNncUUOhAbD0t1KG07GLsMwZ9vyh4ObvffFyPFH/6SXDzpulLmn5Ze93YXp8I/aSo/9rJya3ujXpwTts0xEmMbjVxPCCMhzh7Cc5/UKbN8Z4LHtDknCPsSqLTdUkcCkarAuc0q01BYaFoGoZJSBZHiFDuPuB4OKLpRaik/4OnsxGjccJ22/DBh2doI3B1SVk3HB3tMZwOGIwzHj09J00j4jRiuy05vncXKS111XTKChbdao6PpzSt5f6TBU/OV1gRUDWWp8vCc36NZJDGjIcZkXLEoSKKlP8ZrSYKU7S1NK1jmITsjVI+PN9ysa64M5mgHay3JdkwZbXYUjcGs8iZDFNabVmutkRJxP405cHTS549v6KsWhblioIEiyWIAm+91ClYWOfJIN4mSeymZcDOQ8wpuXMQcc4L8mpj+N6f/DHnF+f8+r/3W5zcPqFtWtq2RXeMKC//Yq4ZWsYg5XU8XatneBVrC6guUfkdwOfI9FIIz03k4zP1PoBfllDbbVtv/P/Lr7nJkOq/J6Uf6b38u5IkZn//wN8oQYhzJbrTxXHWsdluSLIh0WBEm68QCJI0Qa69Mq/VPlA1ju++f8rXXt3D1FtOFxX7s5SqLHkjyYhTQxh49N9qXZCmCQJLmqXUdU0YCMbDBJQC5ciSmDg2PH2+4scfPGVvf8TR8ZTDoxlxEtEYQ141BJGiLTQHhweEoaIoaqq6xWjfR+Tb0h/bUpKXNVGs2J+kVLVhlCbEXXMrpWCQRmRRQJoEPgjbFulioiBEN55VdLXJEcoSx0MenG4pG0PVOp6drnnz1Rnjccp8vvHI0KJhk1fgLPfuHHD37jFOKpbzFVZYpgcHGC05mKX8+KzYWe/Qrf610UBnuNFNYm5eVyW8xZB3He8QZViauuDDH/0Vz5885dd+49f52te/wXA0omn1bmrnHctf5FD3P7ff8vdEk35b5PoR6OfB3gRBsEMv3mw8+8D86SXMizdIPzeHm5n85ZFlf9PATc36o6Nj9vf3sM4RxxFFIXeKaVEUsV4sme3PGEz2WVU5bVORJjGDQcpyW3jckPNCrKebih/cv+Dt4xGmahnYlHgwojYS4xS207msG0NRrknTBEvlb0ihKGtN27QMswRdac7OF/z4/im1htlsynicMhhk5EVJqx2j0YTBOOXy4pxs4GmH1viGbLvOCSLFeJwxnY1ACobLmqzSBIHCDCyj1AtQFXVL22qSAAaJJMtCwiRB6xDhPIlifrFgMhxyuH+IcZrFpmAUK06vSjaVIQQu5xV3j6dEUU3b2N1nKhws1yV52dC2Dft7U/KyROID92g64v1nRZdoOhZax2oS0mvX9AezFBKLxQmPjg0669SbCdFYjbOGstjyB7//+7z3lz/kW3/jO7zy+uskabrjW5seyHZjl9M/emVlutG1gK6ZFZ2K3mcM+uFwSDWvd8FK/8PFi6XIT18y9YH8Irb+5pb2OvCBXSkld6fIF956myzNiJKQotxweXWJa1vKuiZJE+bzOYPRkMn0AF3nrK/OEAIO9obkZUlb+7reywY6rFJY5Y0bUCHpIGGdt2SJt+qMFGgDbWPR1FgJcRSRhDFBKFESTs8uGSUpi5VXLB4PMwbDhCyLyYaJt8VsK+I4JIpCDg73UKGiaWu0tRTbmro2FJUmigI2m4JtZSlqTRqFRM6RxAFpHBKGAVmiiAKJEg7vyRQSRAEuDjHaS5+PBhnDLCZLFVk24mg6YjpY81645MHzDaFSXG5bwmDD3njM6cUc0frtdZpE1E1N03l0yVAwiUdeYcJZos7sTqreVK6fjzfESbJjyVkcKAH2WqtISfHCCdBf/6ZtGCQZOHj67An//H/5J2TpgOneHvtHRxwenTCeTIjjhNFoSBB5M5Ak62iH1uC0IS8LlosF2/WGIs+5OL9gvrjiv/4vfvczBv0gZr7wNoq7oP0p05qXa/2Xx5Qvv+lPaoZvZn0hBOPxmKOjI955522sNdRNxbMnT2nalqZumEzGRGHI1dUV2XBEPN4jLiuaco2ScPtowsNnV9i2cyTRjvN1hXOOW+OIIPLWL1jD80tDVUXMxhFGG7abgsPZCF20jERIdbVknCYo53zDFQYYAbPjPfJlwXy+YZDFLBZb8tyLN5VFRRD5Mdtm02G92xZrBUEQYa1mNhuxLWrevf+E8WyGCDwdbjRMiKIA3WrG45Rx5hlcRVkwm45pG+29cUVDFCvGScpwEJPEsW//hW94Xz3yBOzHpxvWhSIUFdOxxxWVVUNVN1htGI8ThHOUVYOSijRLuFpscDJgOooJpC9TvBKBpK8gXCdh3icyJ70RRw8Q60uU68ToX6d1S9s2RHGCwSKEo6w2NKcVp6dPEaJXmvbJNgoCkigmG46wXWzUZcnF5Rl1VXhv2W6M/WmKCD8z6EdZR1zuoTY3Avjl/355wvNyCXOzxPmk5ZQQviXqJpIIIbh79y5pHBOHEVfzOa/ee5Ufv/8++WVOWdYkacvh8Qn3f/ITLs7P2T88YLB3BFcWUxfcvb2HUpJHz66oG4Nyfo5/lTcgBFKWNKV3ymuHCcY0tMYSKUjTjEpbIgc2BaUCitpQ5BtODkdcrJY8vFyTN5bjW/uoKGS5yjEOWu0xIyLQmOUagWA1z2m1Ic9zxtMBcRiTFzXaekO7QZpgtaGpSxCKpmoIPDEYZ0Abb2wnpaBt/FY4DRVpmKKUQEnBYJiiAkVdtjjpVdkmOuHNuyF5pVnnhjAa0HRQ3DCKwFrqzoHxYG9CNuiCf5NzfrUht4qy495q13pPAn+BfO3dGaXZmxOTrtpxzu0kTq4tdPpk6Gjq3LOqdie9Px2CrpmVUkDojTu0binKlu12DVJw9vwx4E8UGXTEcF5Msp8t6Aee2NvXVNxoTm/+8H6V3Gtidd/07//Gnf7y42M3CNc3TxiEvPbaaxwdHdDUFfsHM7TWjMcTLi/nlGVBXMTs7+8znk25vDgjyRKy4QQlBNVqzpe/9lU23/sur92LePL8guW29FcEL4lXW8GysaQWGldStyHrvGQ8iNgbJV05YbEE1HVFEkniwFLVmvVWI1RCYzbEaQxSsSoaD2OWiiQKIQhZFVsSpcjikMEgZb0puLjYMp0ICBRnFyv2Z2MO9kdcriqU6EZ0bUtta7IsJYtTJBLbatIo8gYMAuI43GmyJ4l3iZFSIuKAIHSEyZhyoBmWGv1GwP/x/3zEXz2sCcNjxsOE5WqFtWBaTfG8pqoNt44nKAlOxZSt5DJvWW39NlYGIWkc+SzeTXCUUsggwN7gMfTXto+HMFA7dYv+OdZajG5p24o49v6v/cbVWdvN5v2ORgbK2+70EyMhCOnLJu9r0AHVP3Wc/qlBnwSdt1NP8/spc/rdv/uY/4TfeRNs9vKNc7OJ7T+s2d4++weHqDBGhSHPn19wcLBHGCQMsgHrzZo8zxkMBhwcHlAUOU8ePebw6JjhaEw0mvHd//vPaaoWGQju3T4kXWy4WuUYbdDdh2fBy4EjuFoXKBzagsGhnN/+1Y0hCiR1bbl9NGaxbfno6ZrzTUtVaYbLCm0t2moWeYsDjvfHbJs1kbLsn0wIpL+oo9GAi/mG55drgkChTQPCEinP+SX0Vj9xHCGsIVABQeCpeda2hEoRhcqP84Skdppnz+Zkg5STE+/j1ScoJQRZHJJGinEWcrT3i/zTP/xL/q8fPuVolvKVt46wWvP8dEUYxSyeb1kU3gFERgErLbg/r/j+RxdsakPiIFCi0wfVnpdMv8T0ZJ2dLunuugvCMKBu2o/FhHNQlSVxnF0rud0YYlgc0nkgm1Shpy7iySPWWoRkJ/exw2EKQWs/x5w+CaIu6F+M4pcD/sUm9Xq11H+hVyboZ64ff+zwdYjO+e+1V19nkA3Z29tju80JQq+NkqYDXn/9dd59713qpmG5WrK3v8/B4SEPPrrP6ZOnHN6yzPZmTKOIfHFJtVlgbc3edECaRsxXBXEHSwW/1fNjNbBKsqk1SkniAOpao03B/nSA0S3hssQ6S+MUq21OUTVE0m8tpXJYJzsR2Zw0DDiaZTw6XTAdJsSBRycOhwPWeUPbNHzxC6+y2ZRsCi92JIMAqb20XdhJfoehomkrjNaM96deM8gYWqvBWQ4PJxRF3eGSBGXhye5hHHdu5SFxJHntJOO//Adf54/+7AHvP7hABiHrvGHVCp5frAgQbJqQ6ThgXZU8XpS892TBomh32Vm3rde06VG01tsYtR1tcEcm6eJBSkGgZKdT6nYZf1cltA1NVZIORj75dTes7WAtvTONVAGilwvvkqNS3mgP102LOth6+NIY/f9X0CtuQCJfDtNPyPTeC7R7jeue078RbqAmP3HE2WV/58Flr7zyKvt7+2y33kA3S72NepalqCBlMpkyX8zZrDcEYUQ6GHB0+4Szp885f/4MZzWT6YzZ0W3yZMB6cU5bFUSB42g2BCfQuvEk5yCk1A5BgNCWQAkaC3XZIpA0lUEvKsqyorUKJR0ykBxMMjaBIB2klE1DKv3RLwioakfV1CyLimEcUDaG2SBBScGqqNhsag73J1zMcxbrirzUWAfCWYRSnVO7IS9LwlT5WjcMqOqGKq9RUhJFkiSNUSHEceCDAEtTVxjpkaZpnBIlITiHbipu7w35O7/2JrqtOL1cU7WW0joqbQjDmA/ONySbhkeXW85XJbU2HfjNqxN4g+PrxaF1ruPHym5r2iNjry99FHodf9OZar+w3cdRFlvSdOhvFuu8SgLXydMYryanBJ3/r0GoblYvPKrz03ym/tpBbz+Wtq8fL9f1P3Vk6T7+uo/P+F8Eo92+dYfxaMZstsdmnSMCSZ5vGI2GTCYTcJbRcEzTNCyXC9aLBVIIRqMJ5sRycXrG08ePybcFk9mM4WjEJIpp8g3FZklTF6ANLoiwQpEbi+gmVGkQ4EJvACbwsE9tHLryjVR7tmF/HLN/MGI/jTmYptSN5cnTJZPZGBCUrSVLQ+JAECgoyoq6hmJsGGR+XNoY+OjxBVEUUFadO4uUNHVLlnlyeSAsB+kAaw1xnBAEkrpsqUtNloXESUQQyJ2FfVVVDIYDsmzAerNBVy2HgwwloW21N2rQhjCAr/3cq/zzP36fhxclMo7YNoLNuuB8WeBuLCX7U7qX07NOYK3HxVgpQHWgvpcSF84HNE6QZRFRHPLw8RnwomAv+ElOUeQMhuOd88xNFO/u+f308MaSqo+nmxTVT3v87KDvXJl+dltw/bgO5hdLopc3sy9/vQ92/wYVb7z5JvsHM4qiZDobsVgsGI1GNI3h3sltFqslcZIync5AwGa9ZrVYMp5NGE3GKKW4PDvj8uI5m+2Svf1DssGYdDgjG+/RVFvy1ZK6LDHGYo2nxUkEWw3F2hEpmGYRAZYkDkkChZI+014VhvXTBbNhzJ39AUezhOCdE1qnuJgXLJcF55uaQRIyHYQMkoi8cchNS1nVHB2MUUGA0I75uuZqsWUyGXs9zrLmbpKSb0uE09y+NfEsMSmJkojBIKbKag9FDnpF4Q59aQ3b7ZY0TTiIE6q66lCokjAKvQS61QQS7h4k/J1vf4H/4V/8kA/ON2yqFt1JXPtl0/X18yYJ/ooafA/Uy577axZ0tbgfVfanAvikprXm6++8ysXZJdtaf2wSKARoXQEjL+myK406ATDZ26/a3TZ2NyK98dwdeejzMKdabTuxpI8H8M96iF6+rX8+L0ISbk5trgPfA9HG4zHHx8edq0TAcrkmiVMc0DYtxwczyqogHQ6om5a9wyO0NmzWa9pLzWA8ZDydoIKQ+AVA+ZMAAAmESURBVCphOb/i7NlTknRBNhgwHI0Io5hkckg0Eljd0FRbdJlj2wZrPImjdVBtwRqIAoiUIZW+zq4aCKSg1g0IRVY0RKFCa8N0khKkMR88vPS2OF2tWQlHXmgmWUgY+Tm+EN5YLYpjEI7zhVcZlmHAaJoyTGO0sWhjcc6LWgWDiGwQE0chTV2j65qmbTg8OiFKvJKvCiWmtQShwlrTKYD5zWl/IiRhyFu3R/zHv/42/93v/zuKVmHc9Uj55knsHWg6AonrZLRFdyM48CIjXomgb0T75Gedwxg4GGT8+le/wL/8s3d9/e2ucVtCCIxu0U1FmKQEHV6nfxitCcMQcWOZeXP+3/scXDfDnwOG0LbaNwqfFNji42Cyj/2yPqhfet3LPweuJzivvPIqw+GY/f29zulbYp2hqVuME9ggQIYBe7Mp88WSJM74ytd+kQ/ee5erq0vy1QbbGOIs4/DkNmEUc3l+xma1ZrtesbyKSNKMKEnIBmOCMGIwnhHMDjCtpi4LtK5o6sLLeAhLbRy1cWydn4dbZ4ilZTYIuL9oGCeKYebY5F74aJRFfOnNI9bbCmshLyri0IsUicqSlhYlBWXdsC1aoiggTRJQNY/mGwZZwvEkIE1CGquIDATK293otuHwcEqS+DGmViGiKcnzrVeECPx4taoaoiDaSeVBpzkjnD8lkpTyasU7twf89rde5R//0Yc0+uPX6LoU8TeMV3W2gCYKFE7b3UjbOuv5rd04UQrfyAoEi+Wav/ftr/LjB6d8eHb1sd/jrKGqcmQY7WDJ17P9TlKmi6mbtcTNMqgP/M8V9HXb3KjX6N74T/+BL9fqL9f4n9TIvvDznOTk5DaDbIjRlrKsmE7HrFZr4iRmsdpytVgTJBmz2Yw0PaOuDXEy4t//23+bf/un3+PDnzxiU5Q01voeYLZHkqRs1ivWyzlVWVFVNVIJguCKMExIk4w4iYmTlDCOSIcDnN2jaWp009JUOW1Td+x9j9y0OKrC4awgXGvGqUU6jzU/OspIY0UYDmiMI4gUi8WGUEkipbjaFCghsdpRNoY4iWibzlURwbbSDCOBZItrYzAxtYIokEShoq5bYAuug/dGMc5ZkiRB62uvXq0bmrZhGk13YkrOD/0wHYyjaHK+cHfC4ShiW9cg2GXM/vqILtDY1fUOjMHK61rfqy97cSYBCOuhls5ZjJOcLTaMkoT/5Le+yX/7z/6IddW+kLW9T4H3FfC+Y3JXp9/U1wnD8DqJOvdCwu1/1ufayG7L4gY5l90dfzNgPz6y/Lik3yfV8y8/+j94Mp5x795dyrJkMhmTFyWjyYiyKBmNMi4vL7n7yl3ifMPtk1v85METri6W/MJXvsg/+ke/w//6e/877//4I9Z5TlVVDAcDhtmA8WjCcDRhuVyQb1dUZUndNJRV40eigSROEqIwIElSkiQjiiLiNCUeZgjpXUaaqsQ0NbqpsUaDCKmt5bISKBTCGapHJYmyCGHJIsH+NGE2G4PxKs3OWbZ5SRKnVFWLnAypm4ZhltJcrLnKK+JAMI4TrPYOJaYxWCUIBimr+ZLZ3tjzSp3XDlUywFlDkngN+DzfYKxjsy6oyobhKGOQZQjhhfgr3eLwLoTDJORkf8z9yzNA7kbLN7FVUkhw11Iru8yvAlTX9DtHt3fAzxWFl9iWFhabgtWm4EuvHPHbv/Lz/NM/+QsaY+lVrqVSHXpTd4Ya0S5+bqpp7CY/XVxZ86LU5KfZaX560OfVrvbqf6GUP7useRmacPNx8wb4OAip728cp89P2T/Yo20NSZpS5BUIGA2H5EVFkRdMJj57XV4tWK42vPvuA/7uf/gr/M3f+A6D4YB33/+QxXzJZrmirWriOCGJEo6PblEMRhjTUpYFRVFgWktR5Z0YrHe9S5KEMAy9e14ckw2HhFHciZF6/qZuGmxdUVeFDxLjsEiWjeu09jWhgkfLiix0ZBGkgWBvEDIcTzBtQxgGaCd4umi4yktKLRg5yabSLCtNOkwoyoZJFjEYpBijsUZRVY3Xh2wbotA7qVjX1fFOMJ55+RBjIQojBoOBX2ZpgxYQBCFV3VCULX/67hnvP1l0e4vrqcgLk7gOfuDxLd11Fl1WbRsQDudRN34x1kmFOOl1+/O64fRywSjc4299/R2u1jn/+of3abS+bpqdxbQaqSxS+mla70d2M6hvCgD0E5CXlTc+c9CXlX5hTN/H8MtjyU9b+/bPudkHvIzNATBWs1guaK32KsVNQ5x4dTNkyHpTcOvkmGfPz5ndPsGONa+8cofix/d58Ogx7/34Lr/wlbcw1jKdjCmKgsePnvPw0RPW6xVhFBJHMWEYcHx8RBSHXM4XFMWWfLtBtw1N3dI2DWVZUnRqwkGgPKE5SYjiiDBMvHKClESjCclw4qXlTEtdlbumDA1WQO0cdeOYV55uF68sg7hmmiqGyZj7S8P37y88jzQJyRtNEkpO1zWtcczSkDR0lHWNaWuqCuqmxGhNUxUMh0PiJP3/2juX3qaBKAp/41ccx2kSlLJGQkKiAglRIVSxZssf4P+yAqGyYQHhIaqyaZvH5DHjx5jF2FaboqjrZr5NpGSTxXU8Pjn3HKJOgBcKlGp86T5pmhL4tnVPrhVV4Nd24Dom2wtABHW5mR2u7R0Jez636khQR3VQ2fS4LNdt15gxpv0zqWkPsX53gTKGv1dzHo0PiDsh714fMZUrPn8/pxLWEky99AOmTcfbPiU0d5/mtVkOp5E0qdWnHewc+tl6fUN93H6q31ZhrvO/965/dht7op1MJpycvEEpzXA0YDqdMhoNkGtFFMdIueLheMTV1SWHh2NMZZ2Bkx9nfPz0hRfPH/P06AlxHLJaLnl5/IzfP884Pf3Krz9nzBYzoGIhF/T7A/wo5GAwZPRgZN2VyxWZ1mi1QasNSimyLGOxWCClJIhCPOGRxDFhJ6ITxYRhRJqmJL2UpJdAvUm0lNI2mBc5mdYIUWKKElUK1EZwocAXPlVZEiUpwtg0trKKmK8KtIassOtxibYKiSkKkm4EXgelNKY0zGZzOnHOwaBPt5eQ5zmbjb1we0nKXEqb1pb0KbDFD/aisI3t3dAn7UZcLrNWDbl1l6Zqk+Waysvm4VWI2zYThI1ar6oKDxu6dTFfonWOR8WwF/P+7Ss2+gPfzmfge9Yp2QobxnYuCO+Gnb39XrVS1FrehV1kuQviLr/SDsd9YrdJweG4h7ihd+wdbugde4cbesfe4YbesXe4oXfsHf8A05ms9YpQUoAAAAAASUVORK5CYII=\n",
|
|
"text/plain": [
|
|
"<Figure size 432x288 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {
|
|
"needs_background": "light"
|
|
},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"import matplotlib.pyplot as plt\n",
|
|
"import matplotlib\n",
|
|
"%matplotlib inline\n",
|
|
"\n",
|
|
"image = cv.imread(\"img/soyjaks.jpg\", cv.IMREAD_COLOR)\n",
|
|
"image2 = cv.cvtColor(image, cv.COLOR_BGR2RGB)\n",
|
|
"cv.circle(image2, (320,220), 55, (143,0,255), 10)\n",
|
|
"cv.putText(image2, 'OMG not chikenz!', (10,90), cv.FONT_HERSHEY_SIMPLEX, 2, (255,0,0), 7, cv.LINE_AA)\n",
|
|
"\n",
|
|
"plt.axis('off')\n",
|
|
"plt.imshow(image2)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "279f1a5d",
|
|
"metadata": {
|
|
"id": "279f1a5d"
|
|
},
|
|
"source": [
|
|
""
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "288aa117",
|
|
"metadata": {
|
|
"id": "288aa117"
|
|
},
|
|
"source": [
|
|
"## Ćwiczenie 2\n",
|
|
"\n",
|
|
"Załaduj obrazy `img/pipe.png` oraz `img/man-without-pipe.png` i wykonaj operacje tak, aby uzyskać poniższy obraz."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 78,
|
|
"id": "2cb83db1",
|
|
"metadata": {
|
|
"colab": {
|
|
"base_uri": "https://localhost:8080/",
|
|
"height": 265
|
|
},
|
|
"executionInfo": {
|
|
"elapsed": 648,
|
|
"status": "ok",
|
|
"timestamp": 1666012106380,
|
|
"user": {
|
|
"displayName": "Cezary Gałązkiewicz",
|
|
"userId": "01409497901784152256"
|
|
},
|
|
"user_tz": -120
|
|
},
|
|
"id": "2cb83db1",
|
|
"outputId": "c3ad7496-2278-40a0-9a51-336108bffe6a"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"<matplotlib.image.AxesImage at 0x7fd69dfbb3d0>"
|
|
]
|
|
},
|
|
"execution_count": 78,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
},
|
|
{
|
|
"data": {
|
|
"image/png": "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\n",
|
|
"text/plain": [
|
|
"<Figure size 432x288 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {
|
|
"needs_background": "light"
|
|
},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"image1 = cv.imread(\"img/pipe.png\", cv.IMREAD_UNCHANGED)\n",
|
|
"image_pipe = cv.cvtColor(image1, cv.COLOR_BGRA2RGBA)\n",
|
|
"image_pipe = image_flipped = cv.flip(image_pipe, 1)\n",
|
|
"image_pipe = cv.resize(image_pipe, None, fx=0.5, fy=0.5, interpolation=cv.INTER_CUBIC)\n",
|
|
"\n",
|
|
"image2 = cv.imread(\"img/man-without-pipe.png\", cv.IMREAD_COLOR)\n",
|
|
"image_man = cv.cvtColor(image2, cv.COLOR_BGR2RGB)\n",
|
|
"\n",
|
|
"background_width = image_man.shape[1]\n",
|
|
"background_height = image_man.shape[0]\n",
|
|
"x, y = 90, 280\n",
|
|
"img_h, img_w = image_pipe.shape[0], image_pipe.shape[1]\n",
|
|
"\n",
|
|
"if x + img_w > background_width:\n",
|
|
" w = background_width - x\n",
|
|
" image_pipe = image_pipe[:, :w]\n",
|
|
"\n",
|
|
"if y + img_h > background_height:\n",
|
|
" h = background_height - y\n",
|
|
" image_pipe = image_pipe[:h]\n",
|
|
"\n",
|
|
"image_pipe_not_trans = image_pipe[..., :3]\n",
|
|
"alpha_mask = image_pipe[..., 3:] / 255.0\n",
|
|
"\n",
|
|
"image_man[y:y+img_h, x:x+img_w] = (1.0 - alpha_mask) * image_man[y:y+img_h, x:x+img_w] + alpha_mask * image_pipe_not_trans\n",
|
|
"\n",
|
|
"\n",
|
|
"plt.axis('off')\n",
|
|
"plt.imshow(image_pipe)\n",
|
|
"plt.imshow(image_man)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "7baba75b",
|
|
"metadata": {
|
|
"id": "7baba75b"
|
|
},
|
|
"source": [
|
|
""
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "f0b793fc",
|
|
"metadata": {
|
|
"id": "f0b793fc"
|
|
},
|
|
"source": [
|
|
"## Co dalej?\n",
|
|
"\n",
|
|
"Przygotuj środowisko pracy instalując wygodne dla Ciebie IDE, np. [PyCharm](https://www.jetbrains.com/pycharm/) (są darmowe studenckie licencje), [Visual Studio](https://visualstudio.microsoft.com) z dodatkiem [Python Tools](https://microsoft.github.io/PTVS/) lub [Visual Studio Code](https://code.visualstudio.com/). Sprawdź czy możesz zaimportować OpenCV, ew. stwórz wirtualne środowisko i doinstaluj niezbędne paczki."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"id": "4a713f31",
|
|
"metadata": {
|
|
"id": "4a713f31"
|
|
},
|
|
"source": [
|
|
"----\n",
|
|
"\n",
|
|
"Źródło:\n",
|
|
"\n",
|
|
"* OpenCV, [Getting Started with Images](https://docs.opencv.org/4.5.3/dc/d2e/tutorial_py_image_display.html).\n",
|
|
"* OpenCV, [Drawing Functions in OpenCV](https://docs.opencv.org/4.5.3/dc/da5/tutorial_py_drawing_functions.html).\n",
|
|
"* P. Pandey, [Face Detection with Python using OpenCV](https://www.datacamp.com/community/tutorials/face-detection-python-opencv)."
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"author": "Andrzej Wójtowicz",
|
|
"colab": {
|
|
"collapsed_sections": [
|
|
"185a63ad",
|
|
"c561f1ea",
|
|
"54e22dd0",
|
|
"2b23ae8a",
|
|
"bb1a0101",
|
|
"cdcb4b70",
|
|
"60678ca6",
|
|
"1219997d",
|
|
"cdba71d3",
|
|
"f0b793fc"
|
|
],
|
|
"provenance": []
|
|
},
|
|
"email": "andre@amu.edu.pl",
|
|
"kernelspec": {
|
|
"display_name": "Python 3 (ipykernel)",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"lang": "pl",
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.8.10"
|
|
},
|
|
"subtitle": "01. Wprowadzenie do widzenia komputerowego [laboratoria]",
|
|
"title": "Widzenie komputerowe",
|
|
"year": "2021"
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 5
|
|
}
|