Effective Classification Of Plant Disease Using Image Processing And Machine Learning

In agricultural domain early identify leaf disease from plant leaf primarily is very crucial for Rapid growth of plants. So prevention of losses of crops are very much depended on appropriate plant disease identification and prevention. So uses of capturing the photos for predicting plant diseases effectively. Here SVM based machine learning approach are applied on area of affected region and classify properly from disease affected image. The work is sub divided into some sections as image acquisition, image pre-processing, feature extraction, segmentation and classification. In pre-process parts are transferred into rgb image with filtering is applied to reduction of noise and after that feature are extracted from images. Next the segmentation is applied for detecting the region of clusters with kmeans approach and other methods as threshold based segmentation or region growing is used to efficiently detect disease affected area. Then Classifier are used for detecting healthy and un healthy images. Here MKSVM based classifier is used for classifying the diseases as Cercospora Leaf Spot, Alternaria Alternata, Anthracnose, Bacterial Blight etc. So some of the statistical parameters are like mean, sd, rms, entropy are measured that are significantly gives good performance for identify disease. Here the results obtained using this approach gives more than 95% of accuracy rates comparing to other approaches of detecting plant diseases efficiently.