API Documentation

Simple FastAPI Backend

The API is built with FastAPI and serves the trained TensorFlow model.

Setup

The API loads a pre-trained model and serves it through two endpoints:

MODEL = tf.keras.models.load_model("../models/1/potatoes.h5")
CLASS_NAMES = ["Early Blight", "Late Blight", "Healthy"]

Endpoints

Health Check

GET /ping

Returns "Hello" to verify the API is running.

Image Prediction

POST /predict

Uploads an image file and returns prediction.

Response:

{
  "class": "Early Blight",
  "confidence": 0.95
}

Dependencies

From requirements.txt: - tensorflow - fastapi - uvicorn - python-multipart - pillow - numpy

Example Usage:

cURL

curl -X POST "http://localhost:8000/predict" \
  -H "accept: application/json" \
  -H "Content-Type: multipart/form-data" \
  -F "file=@potato_leaf.jpg"

Python

import requests

url = "http://localhost:8000/predict"
with open("potato_leaf.jpg", "rb") as f:
    files = {"file": f}
    response = requests.post(url, files=files)
    result = response.json()

print(f"Disease: {result['class']}")
print(f"Confidence: {result['confidence']:.2%}")

JavaScript

const formData = new FormData();
formData.append('file', imageFile);

fetch('http://localhost:8000/predict', {
  method: 'POST',
  body: formData
})
.then(response => response.json())
.then(data => {
  console.log('Disease:', data.class);
  console.log('Confidence:', data.confidence);
});

Response Codes

Code	Description
200	Success - Prediction completed
400	Bad Request - Invalid file format
422	Validation Error - Missing file
500	Internal Server Error

Error Handling

Invalid File Format

{
  "detail": "Invalid file format. Please upload JPG, PNG, or JPEG images only."
}

Missing File

{
  "detail": [
    {
      "loc": ["body", "file"],
      "msg": "field required",
      "type": "value_error.missing"
    }
  ]
}

Rate Limiting

Currently no rate limiting is implemented, but consider implementing it for production use:

• Requests per minute: 60
• Requests per hour: 1000
• File size limit: 10MB

CORS Configuration

The API is configured to accept requests from: - http://localhost:3000 (React development server) - http://localhost:8080 (Alternative frontend port)

Model Information

Input Requirements

• Image Format: JPG, PNG, JPEG
• Processing Size: Images are resized to 256x256 pixels
• Color Space: RGB (3 channels)
• Normalization: Pixel values scaled to [0,1]

Model Details

• Architecture: Convolutional Neural Network (CNN)
• Framework: TensorFlow/Keras
• Model File: ../models/1/potatoes.h5
• Classes: 3 (Early Blight, Late Blight, Healthy)

Integration Examples

React Frontend Integration

// Image upload handler
const handleImageUpload = async (file) => {
  const formData = new FormData();
  formData.append('file', file);
  
  try {
    const response = await fetch('http://localhost:8000/predict', {
      method: 'POST',
      body: formData
    });
    
    const result = await response.json();
    
    setPrediction({
      disease: result.class,
      confidence: (result.confidence * 100).toFixed(1)
    });
  } catch (error) {
    console.error('Prediction failed:', error);
  }
};

Mobile App Integration

// Flutter/Dart example
Future<Map<String, dynamic>> predictDisease(File imageFile) async {
  var request = http.MultipartRequest(
    'POST', 
    Uri.parse('http://localhost:8000/predict')
  );
  
  request.files.add(
    await http.MultipartFile.fromPath('file', imageFile.path)
  );
  
  var response = await request.send();
  var responseData = await response.stream.toBytes();
  var result = json.decode(String.fromCharCodes(responseData));
  
  return result;
}

Performance Metrics

• Average Response Time: ~200ms
• Model Inference Time: ~50ms
• Image Processing Time: ~100ms
• Network Overhead: ~50ms

Security Considerations

WarningProduction Recommendations

• Implement authentication (API keys, JWT tokens)
• Add rate limiting to prevent abuse
• Validate file sizes and types thoroughly
• Use HTTPS in production
• Implement request logging and monitoring

Troubleshooting

Common Issues

CORS Errors: - Ensure frontend runs on allowed origins - Check CORS middleware configuration

Model Loading Errors: - Verify model file exists at correct path - Check TensorFlow version compatibility

Large File Uploads: - Implement file size validation - Consider image compression before upload

Debugging

Enable debug mode for detailed error messages:

# In main.py
app = FastAPI(debug=True)