Rice Leaf Disease Recognition using Gray-Level Co-Occurrence Matrix and Statistical Features

Abstract

Developing a novel technique conducive to plant leaf disease recognition and classification is an expanded research sector. Normally identification of different rice leaf diseases in a rural area is a very difficult task as there is very little scope of connecting with the specialist. So an automated process of identifying rice leaf diseases help farmer from happening ample harm. This proposed scheme can recognize the major occurred five categories of rice leaf diseases (Bacterial leaf blight, Sheath blight, Bacterial leaf blast, Brown spot, and Tungro) by applying a machine learning algorithm. Textural and statistical features are requiring here for extracting features and classifying disease perfectly. 92.77% accuracy achieved in this method and can identify rice leaf diseases satisfactorily in an efficient way. By adding some extra combined features this system can become more helpful for farmers, specialists, and agriculturalists. This will also support agricultural-related people and organizations to identify diseases and take proper action against rice leaf diseases.