PROBABILITY BASED CLASSIFICATION METHOD FOR PLANT DISEASE DETECTION

Nowadays, detecting diseases in plants is a crucial challenge of image processing due to complex nature of the input data. A range of algorithms have been developed by researchers so far to help in this direction. The plant disease detection has the various phases like pre-processing, segmentation, feature extraction and classification. In this work, we have used a technique which uses textural features, then the image is segmented and further classification is performed on the segmented images. The k-mean clustering algorithm is helpful in segmenting the image to form clusters of similar data together. To improve various parametric values like accuracy, precision and recall, we have replaced the pre-existing SVM classifier with the naïve bayes classifier. We have implemented our proposed work using MATLAB. Our results analysis outperforms over existing approaches in terms of accuracy, precision and recall.