Vision Based Detection and Classification of Disease on Rice Crops Using Convolutional Neural Network

This paper presents a combination of K-mean s clustering, SVM and CNN based classification techniques for the crop disease / infection with target as Rice crop. Agricultural crops show variation in features of healthy crops in color, texture and growth. Manual inspection and monitoring of crops is incompatible for large farm sizes so aerial monitoring or camera imaging can prove to be helpful for agriculturists in pre-determination of crop diseases. The use of image processing algorithms involving K-means clustering, Support vector machine and Convolutional Neural Networks enables classification process. The supervised classification techniques involve prior training process which determines the accuracy. The training dataset involved initially with 10 image set which was increased to over 50 to obtain accuracy over 95 percent.