Morphological and Otsu’s Segmentation, Classification and Disease Detection of Maize Plant Using Texture Feature Analysis
Diseases affected on plants causes’ major loss in production as well as economic growth. To enhance crop production it is most important that plant diseases must be analyzed earlier so that effective control actions can be taken almost in advance. This paper discusses about the various fungal diseases of maize, how to identify/classify these diseases using image segmentation along with classification Algorithm. Disease symptoms can be distinguishable as earlier as possible through inspecting either stem or leaf part of the maize. This proposed algorithm/method automatically identifies the maize diseases and classifies whether the stem or its leaf is normal or diseased i.e. having Bacterial, Viral or Fungal disease. Gray Level Co-occurrence Matrix (GLCM) is used for extracting features of infected area. Then diseased leaves of Watershed segmentation or Otsu’s clustered Maize images are classified by Support Vector Machine (SVM). Various fungal maize diseases are taken and can be classified by using different classifiers like Tree, Linear Discriminate, K-Nearest Neighbors (KNN) and SVM.The algorithms can be used for training and classification purpose, out of these, SVM gives better results for most of the application. In this work 40 images of Watershed Segmented and Otsu’s clustering segmented healthy/diseased maize are taken and texture features are calculated using GLCM.Based on these features along with SVM classifier, it can be classified into healthy and diseased with accuracy of high percentage compared with other classifiers.