🌿 Plant Leaf Analyzer - AI-Powered Fertilizer Recommendation App

A web application that uses a TensorFlow Keras model to analyze plant leaf images and provide fertilizer recommendations based on detected diseases.

Features

• AI-Powered Detection: Uses PlantVillage model to classify leaves with confidence scores
• Disease Detection: Identifies Early Blight, Late Blight, and Healthy leaves
• Fertilizer Recommendations: Provides detailed fertilizer suggestions including:
  • Primary treatment with NPK ratios
  • Secondary supplements
  • Application schedule
  • Care tips
• Nutrient Analysis: Visual representation of nutrient levels
• Severity Indicator: Shows the severity level of detected issues
• Responsive Design: Works on desktop and mobile devices

Model Capabilities

The integrated plant_village_model_v1.keras model classifies leaves into:

• Early Blight - Fungal disease (Alternaria solani) with dark concentric spots
• Late Blight - Severe water mold disease (Phytophthora infestans)
• Healthy - Normal, healthy leaf tissue with maintenance recommendations

Installation

Prerequisites

• Python 3.8 or higher
• pip (Python package manager)

Setup Steps

1. Install Python Dependencies

   pip install -r requirements.txt

2. Verify Model File Ensure plant_village_model_v1.keras is in the app directory

Running the Application

Method 1: Using the Batch Script (Windows)

Simply double-click start_server.bat

Method 2: Manual Start

python app.py

The backend server will start on http://localhost:5000

Using the Web Interface

1. Start the backend server (using one of the methods above)
2. Open index.html in your web browser
3. Upload a leaf image (drag & drop or click to browse)
4. Click "Analyze Leaf" button
5. View AI-powered diagnosis with confidence scores and fertilizer recommendations

API Endpoints

POST /predict

Accepts an image file and returns classification results

Request:

• Method: POST
• Content-Type: multipart/form-data
• Body: image file

Response:

{
  "success": true,
  "prediction": "Early_Blight",
  "confidence": 0.95,
  "probabilities": {
    "Early_Blight": 0.95,
    "Late_Blight": 0.03,
    "Healthy": 0.02
  }
}

GET /health

Health check endpoint to verify server status

Technologies Used

• HTML5
• CSS3 (with gradients and animations)
• Vanilla JavaScript
• No external dependencies required

File Structure

Fertilizer App/
├── index.html                      # Frontend interface
├── styles.css                      # Styling and animations
├── script.js                       # Frontend logic with API integration
├── app.py                          # Flask backend server
├── requirements.txt                # Python dependencies
├── plant_village_model_v1.keras   # TensorFlow model
├── start_server.bat               # Windows startup script
└── README.md                       # Documentation

Technology Stack

Frontend:

• HTML5, CSS3, JavaScript
• Fetch API for backend communication

Backend:

• Flask (Python web framework)
• TensorFlow/Keras (ML model inference)
• Flask-CORS (Cross-origin support)
• Pillow (Image processing)

Model:

• TensorFlow Keras model trained on PlantVillage dataset
• Input: 256x256 RGB images
• Output: 3 classes (Early_Blight, Late_Blight, Healthy)

Troubleshooting

Server won't start

• Verify Python is installed: python --version
• Install dependencies: pip install -r requirements.txt
• Check if port 5000 is available

Model prediction fails

• Ensure model file exists and is not corrupted
• Check image format (PNG, JPG supported)
• Verify image is a valid RGB image

CORS errors

• Make sure Flask-CORS is installed
• Verify API_URL in script.js matches server address
• Ensure backend server is running

Future Enhancements

• Support for more plant diseases and crops
• Mobile app version (React Native/Flutter)
• User accounts and analysis history tracking
• Batch image processing
• Multilingual support
• Integration with agricultural databases
• Real-time disease progression monitoring
• Weather-based recommendations

License

Free to use for educational and personal projects.