Compare commits
19 Commits
5662ec7a81
...
dev
Author | SHA1 | Date | |
---|---|---|---|
|
7fe54f35d7 | ||
|
bcf0d04de6 | ||
c1f7cd5d92 | |||
7ac72b91cd | |||
|
030d5e432e | ||
|
80706e2c16 | ||
290772dfa2 | |||
e886e8af7e | |||
5f7e15dbea | |||
|
557e16dfd6 | ||
|
e0046ef98a | ||
|
9f3bbb0d87 | ||
a00c20f3b1 | |||
|
294fb339ed | ||
|
695592b7b6 | ||
|
663ef6d80d | ||
caf98c54cd | |||
|
bf70d8eb9c | ||
68ce008dea |
8
.idea/.gitignore
vendored
@ -1,8 +0,0 @@
|
||||
# Default ignored files
|
||||
/shelf/
|
||||
/workspace.xml
|
||||
# Editor-based HTTP Client requests
|
||||
/httpRequests/
|
||||
# Datasource local storage ignored files
|
||||
/dataSources/
|
||||
/dataSources.local.xml
|
@ -1,6 +0,0 @@
|
||||
<?xml version="1.0" encoding="UTF-8"?>
|
||||
<project version="4">
|
||||
<component name="MarkdownSettingsMigration">
|
||||
<option name="stateVersion" value="1" />
|
||||
</component>
|
||||
</project>
|
11
README.md
Normal file
@ -0,0 +1,11 @@
|
||||
Cat detection
|
||||
|
||||
Quick guide:
|
||||
1. install docker
|
||||
2. build image with `./scripts/build_image.bat`
|
||||
3. run container with `./scripts/start.bat`
|
||||
|
||||
Quick tests guide:
|
||||
1. run unit tests `./scripts/run_tests.bat`
|
||||
|
||||
Have fun!
|
39
basics.ipynb
@ -1,39 +0,0 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 7,
|
||||
"id": "initial_id",
|
||||
"metadata": {
|
||||
"collapsed": true,
|
||||
"ExecuteTime": {
|
||||
"end_time": "2024-01-04T16:38:33.550511800Z",
|
||||
"start_time": "2024-01-04T16:38:33.542353Z"
|
||||
}
|
||||
},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 2
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython2",
|
||||
"version": "2.7.6"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 5
|
||||
}
|
58
cat_detection.py
Normal file
@ -0,0 +1,58 @@
|
||||
from io import BytesIO
|
||||
|
||||
import numpy as np
|
||||
from PIL import Image
|
||||
from keras.src.applications.resnet import preprocess_input, decode_predictions
|
||||
from keras.applications.resnet import ResNet50
|
||||
|
||||
|
||||
"""
|
||||
Recognition file.
|
||||
Model is ResNet50. Pretrained model to image recognition.
|
||||
If model recognize cat then returns response with first ten CAT predictions.
|
||||
If first prediction is not a cat then returns False.
|
||||
If prediction is not a cat (is not within list_of_labels) then skips this prediction.
|
||||
Format of response:
|
||||
{
|
||||
'label': {label}
|
||||
'score': {score}
|
||||
}
|
||||
"""
|
||||
|
||||
|
||||
model = ResNet50(weights='imagenet')
|
||||
|
||||
|
||||
# PRIVATE Preprocess image method
|
||||
def _preprocess_image(image):
|
||||
try:
|
||||
img = Image.open(BytesIO(image.read()))
|
||||
img = img.resize((224, 224))
|
||||
img_array = np.array(img)
|
||||
img_array = np.expand_dims(img_array, axis=0)
|
||||
img_array = preprocess_input(img_array)
|
||||
return img_array
|
||||
except Exception as e:
|
||||
print(f"Error preprocessing image: {e}")
|
||||
return None
|
||||
|
||||
|
||||
# Generate response
|
||||
def _generate_response(decoded_predictions, list_of_labels):
|
||||
results = {}
|
||||
for i, (imagenet_id, label, score) in enumerate(decoded_predictions):
|
||||
if i == 0 and label not in list_of_labels:
|
||||
return None
|
||||
if score < 0.01:
|
||||
break
|
||||
if label in list_of_labels:
|
||||
results[len(results) + 1] = {"label": label, "score": round(float(score), 2)}
|
||||
return results
|
||||
|
||||
|
||||
# Cat detection
|
||||
def detect_cat(image_file, list_of_labels):
|
||||
img_array = _preprocess_image(image_file)
|
||||
prediction = model.predict(img_array)
|
||||
decoded_predictions = decode_predictions(prediction, top=10)[0]
|
||||
return _generate_response(decoded_predictions, list_of_labels)
|
6
docker/.dockerignore
Normal file
@ -0,0 +1,6 @@
|
||||
*.md
|
||||
/venv
|
||||
.git
|
||||
__pycache__
|
||||
.pytest_cache
|
||||
/tests
|
11
docker/Dockerfile
Normal file
@ -0,0 +1,11 @@
|
||||
FROM python:3.11
|
||||
|
||||
WORKDIR /app
|
||||
|
||||
COPY . /app
|
||||
|
||||
RUN pip install --no-cache-dir -r requirements.txt
|
||||
|
||||
EXPOSE 5000
|
||||
|
||||
CMD ["python", "main.py"]
|
10
docker/docker-compose.yml
Normal file
@ -0,0 +1,10 @@
|
||||
version: '3.3'
|
||||
|
||||
services:
|
||||
cat-detection:
|
||||
image: cat-detection
|
||||
build:
|
||||
context: ../
|
||||
dockerfile: ./docker/Dockerfile
|
||||
ports:
|
||||
- "5000:5000"
|
6
docs/basics.md
Normal file
@ -0,0 +1,6 @@
|
||||
#inzynieriaOprogramowania #semestr3
|
||||
|
||||
# Faza 1 - ogólna architektura projektu
|
||||
Aplikacja oparta na micro-frameworku Flask, w której użytkownik przesyła zdjęcie na serwer i w odpowiedzi dostaje na ile % na zdjęciu znajduje się kot (poszczególne przedziały będą miały inną informację typu 100%-95%: "to *prawie* na pewno jest kot" itd.).
|
||||
|
||||
Wyniki oceny będą przechowywane w sesji - będą zapisane do momentu wyłączenia przeglądarki, przez co nie ma potrzeby stawiania bazy danych.
|
53
docs/docs.md
Normal file
@ -0,0 +1,53 @@
|
||||
# Api
|
||||
|
||||
Port -> 5000
|
||||
|
||||
endpoint -> api/v1/detect-cat
|
||||
|
||||
Key -> 'Image'
|
||||
|
||||
Value -> {UPLOADED_FILE}
|
||||
|
||||
Flask Rest API application to cat recognition.
|
||||
If request is valid then send response with results of recognition.
|
||||
If key named 'Image' in body does not occur then returns 400 (BAD REQUEST).
|
||||
Otherwise, returns 200 with results of recognition.
|
||||
Format of response:
|
||||
```json
|
||||
{
|
||||
"lang": "{users_lang}",
|
||||
"results": {
|
||||
"{filename}": {
|
||||
"isCat": "{is_cat}",
|
||||
"results": {
|
||||
"1": "{result}",
|
||||
"2": "{result}",
|
||||
"3": "{result}",
|
||||
"4": "{result}",
|
||||
"5": "{result}",
|
||||
"6": "{result}",
|
||||
"7": "{result}",
|
||||
"8": "{result}",
|
||||
"9": "{result}",
|
||||
"10": "{result}"
|
||||
}
|
||||
}
|
||||
},
|
||||
"errors": [
|
||||
"{error_message}",
|
||||
"{error_message}"
|
||||
]
|
||||
}
|
||||
```
|
||||
Format of result:
|
||||
```json
|
||||
{
|
||||
"label": "{label}",
|
||||
"score": "{score}"
|
||||
}
|
||||
```
|
||||
|
||||
Example response:
|
||||
```json
|
||||
|
||||
```
|
30
language_label_mapper.py
Normal file
@ -0,0 +1,30 @@
|
||||
import os
|
||||
|
||||
from jproperties import Properties
|
||||
|
||||
"""
|
||||
Translator method.
|
||||
If everything fine then returns translated labels.
|
||||
Else throws an Exception and returns untranslated labels.
|
||||
"""
|
||||
|
||||
|
||||
def translate(to_translate, lang):
|
||||
try:
|
||||
config = Properties()
|
||||
script_directory = os.path.dirname(os.path.abspath(__file__))
|
||||
resources_path = os.path.join(script_directory, "./resources")
|
||||
|
||||
# Load properties file for given lang
|
||||
with open(os.path.join(resources_path, f"./{lang}.properties"), 'rb') as config_file:
|
||||
config.load(config_file, encoding='UTF-8')
|
||||
|
||||
# Translate labels for given to_translate dictionary
|
||||
for index, label_info in to_translate.items():
|
||||
label = label_info.get("label")
|
||||
to_translate[index]["label"] = config.get(label).data
|
||||
return to_translate, []
|
||||
except Exception as e:
|
||||
error_message = f"Error translating labels: {e}"
|
||||
print(error_message)
|
||||
return to_translate, error_message
|
153
main.py
@ -1,59 +1,108 @@
|
||||
from PIL import Image
|
||||
import torch
|
||||
import torch.nn.functional as F
|
||||
from torchvision.models.resnet import resnet50, ResNet50_Weights
|
||||
from torchvision.transforms import transforms
|
||||
from flask import Flask, request, Response, json
|
||||
from cat_detection import detect_cat
|
||||
from language_label_mapper import translate
|
||||
from validator import validate
|
||||
from flask_cors import CORS
|
||||
|
||||
# Load the pre-trained model
|
||||
model = resnet50(weights=ResNet50_Weights.DEFAULT)
|
||||
|
||||
model.eval()
|
||||
|
||||
# Define the image transformations
|
||||
preprocess = transforms.Compose([
|
||||
transforms.Resize(256),
|
||||
transforms.CenterCrop(224),
|
||||
transforms.ToTensor(),
|
||||
transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
|
||||
])
|
||||
"""
|
||||
Flask Rest API application to cat recognition.
|
||||
If request is valid then send response with results of recognition.
|
||||
If key named 'Image' in body does not occurred then returns 400 (BAD REQUEST).
|
||||
Otherwise returns 200 with results of recognition.
|
||||
Format of response:
|
||||
{
|
||||
"lang": {users_lang},
|
||||
"results": {
|
||||
{filename}: {
|
||||
"isCat": {is_cat},
|
||||
"results": {
|
||||
"1": {result}
|
||||
"2": {result}
|
||||
"3": {result}
|
||||
...
|
||||
"10" {result}
|
||||
}
|
||||
},
|
||||
...
|
||||
},
|
||||
errors[
|
||||
{error_message},
|
||||
{error_message},
|
||||
...
|
||||
]
|
||||
}
|
||||
To see result format -> cat_detection.py
|
||||
"""
|
||||
|
||||
|
||||
def is_cat(image_path):
|
||||
# Open the image
|
||||
img = Image.open(image_path)
|
||||
# Define flask app
|
||||
app = Flask(__name__)
|
||||
app.secret_key = 'secret_key'
|
||||
CORS(app)
|
||||
|
||||
# Preprocess the image
|
||||
img_t = preprocess(img)
|
||||
batch_t = torch.unsqueeze(img_t, 0)
|
||||
# Available cats
|
||||
list_of_labels = [
|
||||
'lynx',
|
||||
'lion',
|
||||
'tiger',
|
||||
'cheetah',
|
||||
'leopard',
|
||||
'jaguar',
|
||||
'tabby',
|
||||
'Egyptian_cat',
|
||||
'cougar',
|
||||
'Persian_cat',
|
||||
'Siamese_cat',
|
||||
'snow_leopard',
|
||||
'tiger_cat'
|
||||
]
|
||||
|
||||
# Make the prediction
|
||||
out = model(batch_t)
|
||||
|
||||
# Apply softmax to get probabilities
|
||||
probabilities = F.softmax(out, dim=1)
|
||||
|
||||
# Get the maximum predicted class and its probability
|
||||
max_prob, max_class = torch.max(probabilities, dim=1)
|
||||
max_prob = max_prob.item()
|
||||
max_class = max_class.item()
|
||||
|
||||
# Check if the maximum predicted class is within the range 281-285
|
||||
if 281 <= max_class <= 285:
|
||||
return max_class, max_prob
|
||||
else:
|
||||
return max_class, None
|
||||
# Available languages
|
||||
languages = {'pl', 'en'}
|
||||
|
||||
|
||||
image_path = 'wolf.jpg'
|
||||
max_class, max_prob = is_cat(image_path)
|
||||
translator = {
|
||||
281: "tabby cat",
|
||||
282: "tiger cat",
|
||||
283: "persian cat",
|
||||
284: "siamese cat",
|
||||
285: "egyptian cat"
|
||||
}
|
||||
if max_prob is not None:
|
||||
print(f"The image is recognized as '{translator[max_class]}' with a probability of {round(max_prob * 100, 2)}%")
|
||||
else:
|
||||
print(f"The image is not recognized as a class within the range 281-285 ({max_class})")
|
||||
@app.route('/api/v1/detect-cat', methods=['POST'])
|
||||
def upload_file():
|
||||
# Validate request
|
||||
error_messages = validate(request)
|
||||
|
||||
# If any errors occurred, return 400 (BAD REQUEST)
|
||||
if len(error_messages) > 0:
|
||||
errors = json.dumps(
|
||||
{
|
||||
'errors': error_messages
|
||||
}
|
||||
)
|
||||
return Response(errors, status=400, mimetype='application/json')
|
||||
|
||||
# Get files from request
|
||||
files = request.files.getlist('image')
|
||||
|
||||
# Get user's language (Value in header 'Accept-Language'). Default value is English
|
||||
lang = request.accept_languages.best_match(languages, default='en')
|
||||
|
||||
# Define JSON structure for results
|
||||
results = {
|
||||
'lang': lang,
|
||||
'results': {},
|
||||
'errors': []
|
||||
}
|
||||
|
||||
# Generate results
|
||||
for file in files:
|
||||
predictions = detect_cat(file, list_of_labels)
|
||||
if predictions is not None:
|
||||
predictions, error_messages = translate(predictions, lang)
|
||||
results['results'][file.filename] = {
|
||||
'isCat': False if not predictions else True,
|
||||
**({'predictions': predictions} if predictions is not None else {})
|
||||
}
|
||||
if len(error_messages) > 1:
|
||||
results['errors'].append(error_messages)
|
||||
|
||||
# Send response with 200 (Success)
|
||||
return Response(json.dumps(results), status=200, mimetype='application/json')
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
app.run(host='0.0.0.0')
|
||||
|
9
requirements.txt
Normal file
@ -0,0 +1,9 @@
|
||||
flask==3.0.0
|
||||
numpy==1.26.3
|
||||
pillow==10.2.0
|
||||
keras==2.15.0
|
||||
jproperties==2.1.1
|
||||
tensorflow==2.15.0
|
||||
werkzeug==3.0.1
|
||||
pytest==7.4.4
|
||||
flask-cors==4.0.0
|
14
resources/en.properties
Normal file
@ -0,0 +1,14 @@
|
||||
# EN
|
||||
lynx=lynx
|
||||
lion=lion
|
||||
tiger=tiger
|
||||
cheetah=cheetah
|
||||
leopard=leopard
|
||||
jaguar=jaguar
|
||||
tabby=tabby
|
||||
Egyptian_cat=Egyptian cat
|
||||
cougar=cougar
|
||||
Persian_cat=Persian cat
|
||||
Siamese_cat=Siamese cat
|
||||
snow_leopard=snow leopard
|
||||
tiger_cat=tiger cat
|
14
resources/pl.properties
Normal file
@ -0,0 +1,14 @@
|
||||
# PL
|
||||
lynx=ryś
|
||||
lion=lew
|
||||
tiger=tygrys
|
||||
cheetah=gepard
|
||||
leopard=lampart
|
||||
jaguar=jaguar
|
||||
tabby=kot pręgowany
|
||||
Egyptian_cat=kot egipski
|
||||
cougar=puma
|
||||
Persian_cat=kot perski
|
||||
Siamese_cat=kot syjamski
|
||||
snow_leopard=lampart śnieżny
|
||||
tiger_cat=kot tygrysi
|
24
scripts/build_image.bat
Normal file
@ -0,0 +1,24 @@
|
||||
@echo off
|
||||
chcp 65001 >nul
|
||||
|
||||
docker -v >nul 2>&1
|
||||
if %errorlevel% neq 0 (
|
||||
echo [31mDocker is not installed.[0m
|
||||
exit /b 1
|
||||
)
|
||||
|
||||
docker info >nul 2>&1
|
||||
if %errorlevel% neq 0 (
|
||||
echo [31mDocker engine is not running.[0m
|
||||
exit /b 1
|
||||
)
|
||||
|
||||
cd %~dp0
|
||||
echo [32mBuilding docker image...[0m
|
||||
docker-compose -f ../docker/docker-compose.yml build
|
||||
if %errorlevel% neq 0 (
|
||||
echo [31mBuilding docker image failed.[0m
|
||||
exit /b 1
|
||||
)
|
||||
|
||||
echo [32mThe image was built successfully.[0m
|
27
scripts/build_image.sh
Normal file
@ -0,0 +1,27 @@
|
||||
#!/bin/bash
|
||||
|
||||
export LC_ALL=C.UTF-8
|
||||
export LANG=C.UTF-8
|
||||
|
||||
docker -v > /dev/null 2>&1
|
||||
if [ $? -ne 0 ]; then
|
||||
echo -e "\033[31mDocker is not installed.\033[0m"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
docker info > /dev/null 2>&1
|
||||
if [ $? -ne 0 ]; then
|
||||
echo -e "\033[31mDocker engine is not running.\033[0m"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
cd ../docker
|
||||
echo -e "\033[32mBuilding docker image...\033[0m"
|
||||
docker-compose build cat-detection
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo -e "\033[31mBuilding docker image failed.\033[0m"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
echo -e "\033[32mThe image was built successfully.\033[0m"
|
12
scripts/run_tests.bat
Normal file
@ -0,0 +1,12 @@
|
||||
@echo off
|
||||
echo [32mRunning unit tests.[0m
|
||||
|
||||
cd %~dp0
|
||||
pytest ../tests
|
||||
|
||||
if %ERRORLEVEL% equ 0 (
|
||||
echo [32mTests passed successfully.[0m
|
||||
) else (
|
||||
Tests failed.
|
||||
echo [31mTests failed.[0m
|
||||
)
|
9
scripts/start.bat
Normal file
@ -0,0 +1,9 @@
|
||||
@echo off
|
||||
chcp 65001 >nul
|
||||
|
||||
cd %~dp0
|
||||
docker compose -f ../docker/docker-compose.yml up cat-detection -d
|
||||
if %errorlevel% neq 0 (
|
||||
echo [31mStarting docker container failed.[0m
|
||||
exit /b 1
|
||||
)
|
11
scripts/start.sh
Normal file
@ -0,0 +1,11 @@
|
||||
#!/bin/bash
|
||||
|
||||
export LC_ALL=C.UTF-8
|
||||
export LANG=C.UTF-8
|
||||
|
||||
docker-compose -f ../docker/docker-compose.yml up cat-detection -d
|
||||
|
||||
if [ $? -ne 0 ]; then
|
||||
echo -e "\033[31mStarting docker container failed.\033[0m"
|
||||
exit 1
|
||||
fi
|
0
tests/__init__.py
Normal file
BIN
tests/img/Egyptian_cat/Egyptian-Mau-Bronze-750.jpg
Normal file
After Width: | Height: | Size: 360 KiB |
BIN
tests/img/Persian_cat/PER_Fluffy_Fancy_Antony_(5468896948).jpg
Normal file
After Width: | Height: | Size: 100 KiB |
BIN
tests/img/Siamese_cat/siamese-cat-cover.jpg
Normal file
After Width: | Height: | Size: 159 KiB |
BIN
tests/img/cheetah/TheCheethcat.jpg
Normal file
After Width: | Height: | Size: 180 KiB |
BIN
tests/img/cougar/Cougar_at_Cougar_Mountain_Zoological_Park_2.jpg
Normal file
After Width: | Height: | Size: 216 KiB |
BIN
tests/img/jaguar/Junior-Jaguar-Belize-Zoo.jpg
Normal file
After Width: | Height: | Size: 326 KiB |
BIN
tests/img/leopard/African_leopard_male_(cropped).jpg
Normal file
After Width: | Height: | Size: 221 KiB |
BIN
tests/img/lion/Lion_waiting_in_Namibia.jpg
Normal file
After Width: | Height: | Size: 281 KiB |
BIN
tests/img/lynx/552661_poster.jpg
Normal file
After Width: | Height: | Size: 469 KiB |
Before Width: | Height: | Size: 7.1 KiB After Width: | Height: | Size: 7.1 KiB |
BIN
tests/img/snow_lopard/Snow-Leopard.jpg
Normal file
After Width: | Height: | Size: 255 KiB |
BIN
tests/img/tabby/Cat_November_2010-1a.jpg
Normal file
After Width: | Height: | Size: 2.7 MiB |
After Width: | Height: | Size: 345 KiB |
Before Width: | Height: | Size: 105 KiB After Width: | Height: | Size: 105 KiB |
19
tests/test_1.py
Normal file
@ -0,0 +1,19 @@
|
||||
import os
|
||||
|
||||
from werkzeug.datastructures import FileStorage
|
||||
|
||||
from main import app
|
||||
|
||||
|
||||
def test_upload_file():
|
||||
with app.test_client() as test_client:
|
||||
script_directory = os.path.dirname(os.path.abspath(__file__))
|
||||
image_path = os.path.join(script_directory, "./img/tiger_cat/cat1.jpg")
|
||||
image = FileStorage(
|
||||
stream=open(image_path, "rb"),
|
||||
filename="cat1.jpg",
|
||||
content_type="image/jpeg",
|
||||
)
|
||||
|
||||
response = test_client.post('/api/v1/detect-cat', data={'image': image}, content_type='multipart/form-data')
|
||||
assert response.status_code == 200
|
31
validator.py
Normal file
@ -0,0 +1,31 @@
|
||||
"""
|
||||
Validation method.
|
||||
If everything fine then returns empty list.
|
||||
Else returns list of error messages.
|
||||
"""
|
||||
|
||||
# Allowed extensions
|
||||
allowed_extensions = {'jpg', 'jpeg', 'png'}
|
||||
|
||||
|
||||
def validate(request):
|
||||
errors = []
|
||||
try:
|
||||
images = request.files.getlist('image')
|
||||
|
||||
# Case 1 - > request has no 'Image' Key in body
|
||||
if images is None:
|
||||
raise KeyError("'Image' key not found in request.")
|
||||
|
||||
# Case 2 - > if some of the images has no filename
|
||||
if not images or all(img.filename == '' for img in images):
|
||||
raise ValueError("Value of 'Image' key is empty.")
|
||||
|
||||
# Case 3 -> if some of the images has wrong extension
|
||||
for img in images:
|
||||
if not img.filename.lower().endswith(('.png', '.jpg', '.jpeg')):
|
||||
raise ValueError(f"Given file '{img.filename}' has no allowed extension. "
|
||||
f"Allowed extensions: {allowed_extensions}.")
|
||||
except Exception as e:
|
||||
errors.append(e.args[0])
|
||||
return errors
|