Detection and Classification of Coffee Leaf Disease using Deep Learning

Abstract

Ethiopia is the leading coffee exporter in Africa which accounts for 22% of the country’s commodity exports. Coffee is one of the crucial agricultural product in the global economy, particularly for Ethiopia. However, diseases like brown eye spot, wilt, and rust are the most determinant constraints the productivity and quality of coffee export. The disease detection requires specific attention from professionals, which is not achievable for mass production. As a result, an autonomous method for detecting and classifying coffee plant disease become very crucial for better productivity. To determine whether a particular image of a leaf has a brown eye spot, wilt, or rust or if it is healthy, we created a deep learning model trained with image dataset collected from the Wolaita Sodo agricultural research center consisting of 1,120 and augmentation technique also applied to handle data over-fitting problem and totally 3,360 images were used. In order to achieve the best results during the classification of such diseases, we compared training from scratch and transfer learning techniques. Because of this, training from scratch performs at a rate of 98.5%, whereas transfer-based learning offers accuracy rates of 97.01% and 99.89% when employing transfer learning through Mobilnet and Resnet50, respectively. The pre-trained Resnet50 model performs picture classification better than other methods. We are further working towards considering the other class of the Coffee leaf disease by incorporating additional data beside the other pre-trained models.

Topics & Concepts