README.md

KrishokBondhu - Agricultural Assistant Platform

KrishokBondhu is a comprehensive agricultural assistance platform that helps farmers with crop recommendations, fertilizer suggestions, and plant disease detection.

Features

• Plant Disease Detection: Upload plant leaf images to detect diseases
• Fertilizer Recommendation: Get personalized fertilizer recommendations based on soil conditions
• Crop Recommendation: Receive suggestions for suitable crops based on soil and weather conditions
• AI-Powered Chatbot: Agricultural assistance chatbot in Bengali language

Technology Stack

• Frontend: Next.js, TypeScript, Tailwind CSS, Framer Motion
• Backend: FastAPI, Python
• AI/ML: TensorFlow, scikit-learn
• Authentication: Clerk
• Cloud Services: Cloudinary (image storage)

Machine Learning Models

This project uses machine learning models trained on the following datasets:

1. Plant Disease Detection Model

   • Based on Plant Disease Detection Notebook
   • Detects various plant diseases from leaf images
   • Supports multiple crops including tomato, potato, and pepper

2. Fertilizer Recommendation Model

   • Based on Fertilizer Recommendation Notebook
   • Provides fertilizer suggestions based on soil composition
   • Takes into account NPK values and other soil parameters

Data Sources

1. Plant Disease Dataset

   • Based on Plant Disease Detection Notebook
   • Contains images of various plant diseases across different crops

2. Fertilizer Recommendation Dataset

   • Based on Fertilizer Recommendation Notebook
   • Includes soil characteristics and corresponding fertilizer recommendations

3. Bangladesh Soil Data

   • Based on Fertilizer Recommendation Guide 2018
   • Contains area-wise soil composition data for Bangladesh including:
     • Nitrogen (N) levels
     • Phosphorus (P) levels
     • Potasium(K) levels
     • pH values
   • Data integrated into MongoDB for region-specific recommendations

Database Schema

The MongoDB database includes collections for:

• User profiles
• Disease predictions
• Area-specific soil data from FRG 2018
• Fertilizer recommendations
• Crop recommendations

Getting Started

1. Clone the repository:

git clone https://github.com/yourusername/krishokbondhu.git
cd krishokbondhu

2. Install dependencies:

npm install

3. Set up environment variables:

cp .env.example .env.local

4. Run the development server:

npm run dev

5. Start the FastAPI backend:

cd plantdisease_detection_api
pip install -r requirements.txt
uvicorn main:app --reload

API Endpoints

• /disease-detection: Plant disease detection endpoint
• /fertilizer: Fertilizer recommendation endpoint
• /crop: Crop recommendation endpoint
• /message: Chatbot interaction endpoint

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Plant Disease Detection model based on work by babykamal on Kaggle
• Fertilizer Recommendation model based on work by babykamal on Kaggle
• Next.js for the frontend framework
• FastAPI for the backend framework