AI · Telegram Bot · Bangla
Rice Leaf Disease Detection
ধান রোগ নির্ণয় বট
An AI-powered Telegram bot that instantly identifies rice leaf diseases from a photo and delivers treatment advice in Bangla — built for Bangladeshi farmers.
Telegram
MIT License
বাংলা
Python 3.14+
Try the bot
Scan to open on Telegram
Farmers can scan it directly to start chatting with the bot.
@Ricedisease_bd_bot
How it works
1
Take a photo
Place the rice leaf on white paper and photograph it in good lighting
2
Send to the bot
Open Telegram, find the bot, and send the photo directly
3
AI analyses the image
A trained computer vision model classifies the disease with a confidence score
4
Receive advice in Bangla
Google Gemini generates full treatment guidance — medicines, steps, and prevention
Diseases detected
🔴
Bacterial Leaf Blight
ব্যাকটেরিয়াল লিফ ব্লাইট
🟤
Brown Spot
ব্রাউন স্পট
💥
Leaf Blast
লিফ ব্লাস্ট
🟡
Tungro
টুংরো
Model
🤗
Rice Leaf Disease Classifier
ONNX · CPU inference · hosted on Hugging Face Hub
Run it yourself
1
Clone & install
https://github.com/nihal4/Rice-Leaf-Disease-Detection.git
cd rice-leaf-disease-bot
pip install -r requirements.txt
2
Set environment variables
| Variable | Description |
|---|---|
| TELEGRAM_BOT_TOKEN | From @BotFather on Telegram |
| GEMINI_API_KEY | From Google AI Studio |
| WEBHOOK_URL | Your public HTTPS deployment URL |
| WEBHOOK_SECRET | Random secret, min 16 characters |
| HF_MODEL_URL | Direct link to .onnx model on Hugging Face |
| CLASS_INDICES_PATH | Path to class_indices.json |
3
Start the bot
python bot.py
Built with
Telegram Bot API
Chat interface
Google Gemini 2.0
Bangla advice
TensorFlow / Keras
Model training
ONNX Runtime
Fast CPU inference
Hugging Face
Model hosting
License
MIT License
Free to use, modify, and distribute. See LICENSE file for details.