Detecting Tomato Crop Diseases with AI: Leaf Segmentation and Analysis
Ankita Gangwar, Geeta Rani, Vijay Pal Dhaka, Sonam Sonam
Abstract
Plant diseases are a major concern for tomato growers, often causing significant crop losses of around 50 to 60%. Early detection and reliable disease management strategies are essential to reduce these losses. Correctly detecting diseases in tomato crop may prove important for reducing the crop loss. Various researchers focused on this domain and applied deep learning techniques for disease detection. However, the techniques proposed in literature reports high accuracy of detection, faces with the issue of low reliability. This study proposes a deep learning-based approach for tomato disease detection using image segmentation. The authors use VIA tool for creating masks of leaves. The proposed method utilizes customized U-Net model for segmentation followed by convolutional network for classifying the segmented images to ten categories. The accuracy of 98.12% prove that it is a promising technique for automated tomato disease detection, which can contribute to improving tomato production and reducing crop loss.