I built a flexible, open-source computer vision toolfor plant disease detection. My work focuses on using SVM classifiers and texture features from leaf images to identify tomato diseases efficiently and reliably. It’s designed for low-cost agricultural environments and can be extended into mobile-based greenhouse monitoring.
- Goal: Detect tomato leaf diseases from images.
- Method: SVM binary classification using GLCM-based texture features.
- Data: PlantVillage tomato leaf images
- Value: Open-source, modular, efficient, and adaptable for mobile deployment.
This project uses a fork of the scimage project which has a modified GLCM function. To see the details of this modification, please refer to my fork here