Disease Detection and Diagnosis on the Leaves using Image Processing
D Anil, M N Sindhushree, M V Tejaswini, Tehleel Ahmad Lone
Abstract
Agriculture is the backbone of the Indian economy. Almost 65 % of people rely upon it and share a major part of the GDP. Disease detection plays a vital role in prevention of quantity and quality of the agricultural product. Leaf shows the symptoms by changing color and showing the spots on it. Earlier methodology was manual observation of leaves for diseases that took time and were costly. This paper proposes a software based solution to automatically detect and classify plant leaf diseases by image processing of leaf images. The main aim is to detect the disease in the leaf at an early stage and thus take suitable measures to stop this disease. Features are then extracted from the processed image for classification using tensorflow object detection model. The farmer after uploading an image can click on the predict option which first previews the uploaded image and then predicts the disease type of the crop and outputs the details of disease type, description of the disease, symptoms and the treatment or required amount of fertilization for the diseased leaf. This approach will enhance productivity of crops. The experiments have been performed on the diseased leaf images of grape, potato and tomato of the Plant Village dataset.