Detection and classification of grape leaf diseases using deep learning

Abstract

Agriculture attains significant consideration in India due to the rapid population newlineexplosion and increased food scarcity. The Grape is the widely cultivated fruit crops of India as newlineit effectively grow under tropical condition. The Grapes are demonstrated as the profitable and newlinethe cost effective crops of India. However various diseases hinder the cultivation of grapes and newlinecreate huge economics loss to the farmers. The prior identification of disease help the farmers to newlinetake necessary action tom prevents the crop from the disease. Further, the severity of the disease newlinehelps to take decision on the proper usage of the pesticides. The early detection with elevated newlineaccuracy is the crucial step required for the enhancement in the agricultural production. newlineTraditionally the grape plant disease is identified by the naked eye observation of the agricultural newlineexperts. However, the traditional method is impractical due to lack of expert, expensive and time newlineconsuming. The image processing technique attains significant consideration among the experts newlinein the area of disease identification. The pest attack is easily identified by the image of the plant. newlineYet the quality get degraded by the unwanted distortion or noise that degrades the prediction newlineaccuracy of the system. For accurate identification of the disease the image has to go through newlinevarious stages prior to the classification. The Artificial intelligence integrated with the image newlineprocessing is proved to be effective in the leaf disease identification process. This research newlinehighlights the segmentation and classification process for the accurate prediction of plant newlinedisease. Hence, this research introduces a novel Adaptive snake approach for the grape leaf newlinesegmentation. The performance evaluation is carried using the plant leaf data set based on recall newlineand precision. The image classification process is established by proposing the CNNC and IKNN newlinemodel. The analysis is done to prove the efficiency of the proposed CNNC and IKNN model. newlineThe performance metrics su