Waste Classifier API

This Flask API provides image classification for waste types using a trained deep learning model. The API can classify waste images into two categories: Organic and Recyclable.

🚀 Features

• Image Classification: Upload waste images and get predictions with confidence scores
• Health Monitoring: Check API status and model availability
• Class Information: Retrieve supported waste categories
• Cross-Origin Support: CORS enabled for web applications
• Error Handling: Comprehensive error responses and logging

📋 API Endpoints

POST /predict

Classify a waste image from Base64 encoded data.

Request Body:

{
  "image": "data:image/jpeg;base64,/9j/4AAQSkZJRgABAQAAAQ..."
}

Response:

{
  "success": true,
  "result": {
    "class": "Recyclable",
    "confidence": 0.95,
    "all_predictions": {
      "Organic": 0.05,
      "Recyclable": 0.95
    }
  }
}

GET /health

Check API health status and model availability.

Response:

{
  "status": "healthy",
  "model_loaded": true
}

GET /classes

Get list of supported waste classes.

Response:

{
  "classes": ["Organic", "Recyclable"]
}

🛠️ Installation

1. Clone the repository:

   git clone <repository-url>
   cd waste_classification_api

2. Install dependencies:

   pip install flask flask-cors tensorflow numpy opencv-python pillow

3. Ensure model file exists: Make sure waste_model.keras is in the project root directory.

4. Run the application:

   python app.py

The API will be available at http://localhost:5000

📦 Dependencies

• Flask: Web framework for the API
• Flask-CORS: Cross-Origin Resource Sharing support
• TensorFlow: Deep learning framework for model inference
• NumPy: Numerical computing for array operations
• OpenCV: Computer vision library for image processing
• Pillow: Python Imaging Library for image handling

🔧 Model Information

• Input Size: 224x224 pixels
• Categories: Organic, Recyclable
• Format: Keras model (.keras file)
• Preprocessing: RGB conversion, resizing, normalization (0-1 range)

📝 Usage Examples

Using cURL:

curl -X POST http://localhost:5000/predict \
  -H "Content-Type: application/json" \
  -d '{"image": "data:image/jpeg;base64,<your-base64-image>"}'

Using Python requests:

import requests
import base64

# Read and encode image
with open('waste_image.jpg', 'rb') as f:
    image_data = base64.b64encode(f.read()).decode()

# Make prediction
response = requests.post('http://localhost:5000/predict', 
                        json={'image': f'data:image/jpeg;base64,{image_data}'})
result = response.json()
print(f"Prediction: {result['result']['class']} ({result['result']['confidence']:.2f})")

⚙️ Configuration

The API runs on:

• Host: 0.0.0.0 (all interfaces)
• Port: 5000
• Debug: Disabled for production

🚨 Error Handling

The API returns appropriate HTTP status codes:

• 200: Success
• 400: Bad Request (missing image data)
• 404: Endpoint not found
• 500: Internal server error (model issues)

📊 Supported Image Formats

• JPEG
• PNG
• Other PIL-supported formats
• Base64 encoded images

🔍 Troubleshooting

1. Model not loading: Ensure waste_model.keras exists in the project directory
2. Import errors: Install all required dependencies
3. Image processing errors: Check image format and Base64 encoding
4. CORS issues: Flask-CORS is enabled by default