PLANT DISEASE DETECTION USING FEATURE EXTRACTION AND ENSEMBLE CLASSIFICATION

The major issue encountered in image processing is the detection of plant disease. There are four stages in which plant disease detection is carried out namely pre-processing, segmentation, feature extraction and classification. In the pre-processing phase, the image is made noise-free or its contrast is enhanced. Region based segmentation is a widely adopted approach for identifying disease affecting the plant. To extract features, texture feature analysis algorithm is used. Texture features are mined with the implementation of GLCM (Gray Level Co-occurrence Matrix). In the classification phase, classifiers are deployed to predict diseases according to the mined properties. SVM (Support Vector Machine), KNN, DT (Decision Tree) and RF (Random Forest) are the widely applied algorithmic schemes. Several algorithms are developed based on these 4 steps for plant disease detection. The performance of the previously developed algorithm is calculated with respect to various metrics like accuracy, recall and so on.