A Flutter-based mobile application for detecting rice leaf diseases using a TensorFlow Lite (TFLite) model.
- 📷 Select rice leaf image from Gallery or Camera
- 🤖 Run deep learning inference using TFLite model
- 🔍 Displays disease classification with confidence score
- 📡 Live Camera Detection using device's camera
- ✅ Lightweight & works offline
- Flutter
- TensorFlow Lite
- camera & image_picker
- TFLite model trained on rice leaf disease dataset
The app can identify the following rice leaf conditions:
| Disease | Description |
|---|---|
| Bacterial Leaf Blight | Bacterial infection causing leaf wilting |
| Brown Spot | Fungal disease with brown lesions |
| Healthy | Normal, disease-free leaves |
| Leaf Blast | Fungal infection affecting leaf blades |
| Leaf Scald | Bacterial disease causing scalded appearance |
| Narrow Brown Spot | Fungal disease with narrow brown lesions |
| Neck Blast | Fungal infection at the neck of rice panicles |
| Rice Hispa | Insect damage creating characteristic patterns |
| Sheath Blight | Fungal disease affecting leaf sheaths |
| Tungro | Viral disease transmitted by leafhoppers |
- Launch the app on your mobile device
- Choose input method:
- Tap "Camera" to capture a new image
- Tap "Gallery" to select an existing image
- Use "Live Detection" for real-time scanning
- Point or select a rice leaf image
- View results with disease classification and confidence score
- Take action based on the detected disease type
https://colab.research.google.com/drive/1TpPNVuus9liP2qMGzTFFY6cBOTc4xIpH?usp=sharing
Thanks to BIMA (Ministry of Education and Science) for funding this research under contract number B/003/UN46.1/PT.01.03/BIMA/PL/2025