For an Agro-based country like India where agriculture acts as a main source of livelihood for more than 50% of the population, crop diseases are a major threat to food security. Hence, digital image processing along with proper machine learning algorithms can be utilized for the classification of diseases from the images of a plant. In this paper, a comparative study on the effects of different machine learning models on crop disease prediction has been done. Since Convolutional Neural Network (CNN) proved to be the best for image classification techniques, models based on CNN alone were considered in this study. We compared the performance of smallCNN with three pre-trained CNN models namely, AlexNet, ResNet-50, and VGG-16. SmallCNN is the CNN model built by us with fewer parameters and suitable for small datasets. The crop tested in this research is Oryza Sativa (Asian Rice) commonly referred to as paddy which is cultivated in abundance in India. The input dataset was fed into the model after performing appropriate pre-processing techniques followed by segmentation. The best accuracy of 66.67% was achieved in the case of ResNet-50 with Adam as the optimizer at a learning rate of 0.0001.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.