Skip to content

Latest commit

 

History

History
125 lines (100 loc) · 6.69 KB

README.md

File metadata and controls

125 lines (100 loc) · 6.69 KB
Raw

KisanSahayak - Empowering Farmers with Smart Agriculture Solutions

KisanSahayak is a smart, data-driven web application designed to help farmers make informed decisions. By integrating machine learning, environmental data, advanced imaging techniques, and innovative marketplace features, we aim to provide farmers with real-time insights into rainfall patterns, crop recommendations, disease management, and more.

Table of Contents

Project Overview

KisanSahayak offers farmers a comprehensive solution for smarter agricultural practices, utilizing cutting-edge technologies such as machine learning and computer vision. Key highlights include:

  • Rainfall & Climate Analysis: In-depth analysis of district-wise rainfall patterns and climate variations to help farmers plan their sowing and irrigation strategies.
  • Crop Recommendations: AI-based suggestions for optimal crops to grow, based on soil nutrients, temperature, and humidity.
  • Disease Prediction & Management: Our system predicts potential crop diseases and offers management tips, using both traditional and advanced methods like hyperspectral imaging.
  • Voice Assistance: A voice-enabled system that allows farmers to interact with the app hands-free, making it user-friendly for all.
  • Image Analysis for Disease Detection: Farmers can upload photos of crops, and the app selects the best out of 3 images to analyze for disease detection.
  • Multilingual Support: Farmers can access the platform in multiple regional languages, ensuring ease of use for everyone.
  • Trusted User Marketplace: A secure, verified platform for farmers to sell or buy products from other farmers or retailers.
  • Disease Prediction: The app predicts the most likely disease in a farm based on current environmental and crop conditions.

KisanSahayak aims to simplify agricultural decision-making, reduce risks, and improve yields for farmers across India.

Features

  • Rainfall Analysis: Understand actual vs. normal rainfall and deviation trends.
  • Crop Recommendations: AI-based suggestions for ideal crops to grow in specific environmental conditions.
  • Disease Prediction: Forecast potential crop diseases and offer actionable management tips.
  • Interactive Reports: Generate detailed analysis reports from user-uploaded data.
  • Voice Assistance: Voice-controlled features for ease of use.
  • Best Camera Angle Selection: Automated selection of the best image out of three for better accuracy in disease detection.
  • Multilingual Support: Access the app in various regional languages for easy navigation.
  • Trusted User System: Farmers and retailers are verified to maintain trust and transparency in the marketplace.
  • Marketplace: A platform for farmers to trade goods with each other or retailers, ensuring fair access to resources.
  • Most Likely Disease Prediction: AI-powered predictions on the most probable diseases affecting the crops in a specific farm.

Technologies Used

  • Backend: Node JS
  • Frontend: HTML, CSS, JavaScript, React, React Native
  • Machine Learning: scikit-learn, pandas, TensorFlow
  • Database: MongoDB
  • Deployment: AWS EC2, Docker
  • Additional Technologies: NLP for voice recognition, image processing for camera selection, multilingual integration.

Getting Started

To get started with the project, clone the repository and install the necessary dependencies.

git clone https://github.com/yourusername/kisansahayak.git
cd kisansahayak
cd backend
npm install
cd ../frontend
npm install
cd ../ML
pip install -r requirements.txt

Website Usage

Once the project is set up, you can start the web application using the following command on your root directory:

cd frontend
npm run dev
cd ../backend
npm start
cd ../ML
uvicorn app:app --reload

Open your browser and navigate to http://localhost:5173/ to use the application.

App Usage

You can also start the mobile application using the following command on your root directory:

cd android/frontend
npx expo start
cd ../../backend
npm start
cd ../ML
uvicorn app:app --reload

Download Expo Go app on your mobile device

After the app has started running, a QR code will be generated in your terminal or browser. Open the Expo Go app on your mobile phone and use the built-in camera or Expo Go to scan the QR code. Once scanned, the mobile app will load directly in the Expo Go app.

Dataset Information

Our data is sourced from reliable datasets like IMD (India Meteorological Department) and district-wise agricultural reports. The data contains key features like district names, actual rainfall, normal rainfall, percentage deviation from the norm, soil nutrients (NPK), temperature, and humidity.

Future Work

We are actively working on enhancing KisanSahayak by introducing:

  • AI-powered AR Image Assistance: While capturing crop photos, the app will use augmented reality (AR) to guide farmers on the optimal way to take the picture for disease detection.
  • Hyperspectral Reflectance Method: We are exploring hyperspectral imaging techniques to predict crop diseases more accurately, leveraging the reflectance properties of leaves and plants.
  • Expanded Crop and Disease Prediction Models: Building more robust models to predict diseases in a broader range of crops and environmental conditions.

These features are in development and will be added to future versions of KisanSahayak.

Team Members

This project is a collaborative effort by:

  • Sagnik Basak (Team Leader) - Machine Learning Engineer & Data Analyst
  • Tamojit Das – Full Stack Development, Application System Design & Project Manager
  • Ankan Das – DL Model Design & Development
  • Debeshee Sen – Full Stack Development & UI/UX Design
  • Titas Kabiraj – Front End, UI/UX Design & Documentation
  • Ritesh Das – Android Development & Full Stack Development

For inquiries, feel free to contact us via [email protected].

License

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

Note: This is a private repository, and we are not currently open to contributions. However, we will be happy to consider your ideas in the future. Stay tuned!