Diseases Recognition for Oryza Sativa Leaf Plant Based on Artificial and Convolutional Neural Network

B. Karthiga

Abstract: Plant disease is one of the major problems in the agriculture sector. Plants are affected by a various factor such as bacteria, fungi, viruses etc. In this Project, we propose a structure to detect a paddy leaf disease more accurately. The proposed system consists of artificial and convolutional neural network (feed-forward artificial neural network) for plant disease classification. The different types of classifier is analysed. ANN consists of machine learning algorithm and it is trained by choosing feature value that may well classify four type of diseased samples appropriately. CNN architecture includes the three hidden layer (convolution, pooling, fully connected layer) and it used to classify more accurately. As a result, few diseases that usually occurs in paddy plants such as bacterial blight, brown spot, leaf blast and leaf streak are detected. CNN model achieves more accuracy for identifying the leaf disease in the paddy plant thereby showing the feasibility of its usage in real time application.

Keywords: Deep learning, CNN, Paddy plant, Pooling layer