Our Pest and Disease Prediction System is a high-accuracy, AI-powered solution designed to help farmers identify and manage pests and crop diseases effectively. The system achieves a 95% accuracy rate, identifying 8 types of pests and 15 crop diseases. It integrates real-time weather and soil data to provide customized treatment plans and enhance agricultural decision-making. Our model can be used offline, ensuring accessibility even in remote areas.
Plant diseases classes : 'Pepper__bell___Bacterial_spot', 'Potato___Early_blight', 'Potato___Late_blight', 'Tomato_Bacterial_spot', 'Tomato_Early_blight', 'Tomato_Late_blight', 'Tomato_Leaf_Mold', 'Tomato_Septoria_leaf_spot', 'Tomato_Spider_mites_Two_spotted_spider_mite', 'Tomato__Target_Spot', 'Tomato__Tomato_YellowLeaf__Curl_Virus', 'Tomato__Tomato_mosaic_virus',
Pest classes : 'aphids', 'armyworm', 'beetle', 'bollworm', 'grasshopper', 'mites', 'mosquito', 'sawfly', 'stem_borer'
we are working on other datasets ....
Uses a fine-tuned custom MobileNetV2 model, optimized for transfer learning.
Identifies 8 pests and 15 crop diseases with 95% accuracy.
Works offline, enabling farmers to diagnose issues without an internet connection.
Provides AI-generated treatment recommendations based on:
10-day weather forecasts from the OpenWeather API.
Soil pH and moisture data from NASA's API with 99% accuracy.
Gemini API for pest and disease treatment plans.
Dynamically updates based on real-time weather and soil conditions.
Facilitates resource sharing, group discussions, and collaboration.
Encourages knowledge exchange between farmers.
Assists farmers with agricultural queries.
Supports Tamil and Telugu for regional accessibility.
Powered by a third-party chatbot using Zapier.
Provides a comprehensive listing of government schemes for farmers.
Helps farmers access financial aid and support programs.
Designed with HTML and Tailwind CSS for efficiency and responsiveness.
Can be accessed offline, ensuring availability even in remote areas.
HTML, Tailwind CSS
Django Framework
Fine-tuned custom MobileNetV2 (Transfer Learning for pest and disease detection)
OpenWeather API (Weather Forecasts)
NASA API (Soil pH and Moisture Data with 99% accuracy)
Gemini API (Pest and Disease Treatment Plans)
Third-party chatbot using Zapier
git clone https://github.com/Stephen13-05/AgroVision.git
python -m venv venv
source venv/bin/activate # On Windows, use venv\Scripts\activate
pip install -r requirements.txt
python manage.py makemigrations python manage.py migrate python manage.py runserver
Your backend will be live at: http://127.0.0.1:8000/
🔗 API Keys Setup You need API keys for OpenWeather, NASA, and Gemini APIs.
1️⃣ Obtain API Keys
OpenWeather API – For weather forecasts. NASA API – For soil pH & moisture data. Gemini API – For AI-generated treatment plans.
2️⃣ Add API Keys to Environment Variables
Create a .env file in the project root and add:
OPENWEATHER_API_KEY=your_openweather_api_key NASA_API_KEY=your_nasa_api_key GEMINI_API_KEY=your_gemini_api_key
1️⃣ Detect Pests & Diseases 📸 Upload a photo of the affected crop → ✅ AI predicts the issue offline.
2️⃣ Get Treatment Plans 🧪 AI suggests treatment based on weather & soil data.
3️⃣ Check Weather & Soil Data 🌦 Get real-time weather updates & 99% accurate soil pH/moisture data.
4️⃣ Join Farmer Community 👨🌾 Discuss & share knowledge with other farmers.
5️⃣ Use the AI Chatbot 🤖 Ask farming questions in Tamil or Telugu → Get instant answers.
6️⃣ Find Government Schemes 📜 Browse & apply for farming subsidies and support programs.
7️⃣ Offline Web App 📱 Use the app anytime, anywhere – no internet needed!
🌱 Boost crop health & prevent losses with AI-powered farming! 🚜
It is the output of the 36 hr hackathon held at # IIIT Sricity ABHISARG 25