Rice Plant Disease Detection using Multilayer Sparse Convolution Neural Network

Diagnosis of the plant disease has become crucial in agriculture sector due to its importance and application. Disease on any particular region of the plant will expand and propagate on entire regions of the plant. Plant pathology has to be utilized to analyse the plant disease. In order to diagnosis plant diseases using plant pathology, image segmentation techniques have been widely applied. Automatic image segmentation such as watershed method, region based method and thresholding method has been employed in existing by mapping or grouping the continuous and discontinuous features of the plant disease effected regions. Its characteristic of region has been computed using grey level, texture and colour component. Segmented region of the image is employed for feature selector using simulated annealing. For precise plant classification on leaf disease, image segmentation faces several challenges in object detection. To tackle these challenges, an object change detection using deep learning architectures has to be carried. out. In this article, a new plant disease diagnosis method has been proposed as multilayer sparse deep convolution neural network for enhanced plant disease classification.