Design And Development Of Expert System For Diagnosis Of Pests On Grapes Crop

Abstract

India is Agriculture based country, its main occupation is farming. Grapes cultivation in India newlineis produced in large scale. Along with high production rate, reduction in quality is also on high newlinerate, which degrades the quality and quantity of agriculture product. Through literature review newlineit reveal that modern agricultural methods like digital image processing and analysing newlinetechniques are acquired for detection and classification of common pests in crops. Early newlinedetection and treatment of pest will reduce the economic loss. The researcher has done the pest newlinedetection analysis specifically on grapes crops plant leaves. As leaves are mostly infected part newlineof the plant and very easy to capture the pest infected leaf image to diagnose the pest infected newlinewith. The common diseases on grapes crop are Downy mildew, powdery mildew, black rot, newlineleaf blight, esca, anthracnose etc. newlineManual diagnosis of disease causes confusion in accurate detection of diseases due to some newlinesimilarities found when observed with our naked eyes. And continuous monitoring of large newlinefields for pest diagnosis of crops by knowledgeable expert is not possible practically. And also newlineit is not cost effective to appoint the expert person for monitoring. newlineRepeated spraying of pesticides reduces the fertility of soil also it cannot be an all-time solution newlineto get rid of pest infection. To overcome these problems, in many of the earlier reviews newlineresearchers have suggested for early detection of pests without depending on experts for finding newlinesolution, because dependence on experts is not only time consuming but expensive also. In newlineorder to overcome the dependency on expert s knowledge, as reviewed in many of the articles newlinethat, researchers have worked for automation of the system. Which in turn has lead for the newlinerequirement of development of the experts system. newlineFor the development of expert system image processing, Machine learning algorithms and newlinedeep learning technique as Convolution Neural Networks are used.