Real-Time Phytopathology Detection via Deep Learning Rose Leaf Disease Detection App is a sophisticated mobile-first solution designed to bridge the gap between advanced computer vision and practical agriculture. By leveraging a high-performance YOLOv10 architecture, the system provides instant, localized analysis of rose leaf health, enabling early intervention and reducing crop loss.
-
Frontend (Mobile): Built with Flutter. It utilizes the camera and image_picker plugins for real-time streaming and high-resolution captures, ensuring a 60 FPS UI experience across both iOS and Android.
-
Backend (Inference Engine): A Flask(Python) REST API. Flask acts as the orchestration layer, receiving image buffers, preprocessing them, and passing them to the model.
-
Deep Learning Model: YOLOv10 (You Only Look Once). Chosen for its NMS-free (Non-Maximum Suppression) training, which significantly reduces latency and improves accuracy in detecting small-scale disease spots (like early-stage rust or black spot).
-
Data Communication: Secure POST requests with multipart/form-data for image transmission, returning JSON payloads containing bounding box coordinates and confidence scores.
- AI-powered disease detection using YOLOv10 model
- 95% accuracy in classifying rose leaf diseases
- Optimized for mobile inference using TensorFlow Lite (35% faster predictions)
- User-friendly UI/UX with Flutter framework
- Offline prediction support (no internet needed after initial setup)
- Architecture:
- CNN (Convolutional Neural Network) for classification
- YOLOv10 for object detection & localization
- Accuracy Achieved: ~95%
- Frameworks: TensorFlow, PyTorch, Ultralytics, Jupyter Notebook