AI Based Soybean Crop Disease Detection

Abstract

Plant disease affect the growth of crop. Therfore, the earlier identification of disease is very important. This study is based on the use of machine learning in Image processing for the classification of various soybean leaf diseases. This study has proposed a optimized Convolutional Neural Network (CNN)-based technique for soybean leaf classification and detection. CNN is used in back word propagation for training the algorithm in order to improve accuracy and the overall system produces better results with higher accuracy. This research employs a standard data set of soybean leaves as well as real time data collection. ML methods can be applied to detect plant diseases, prior to the full appearance of symptoms. Plant pathologists desire an accurate and reliable soybean plant disease detection system. In this study, we propose an efficient soybean diseases identification method based on a transfer learning approach by using pretrained AlexNet and GoogleNet convolutional neural networks (CNNs). The proposed AlexNet and GoogleNet CNNs were trained using 649 and 550 image samples of diseased and healthy soybean leaves, respectively, to identify three soybean diseases. We used the five-fold cross-validation strategy. The proposed AlexNet and GoogleNet CNN-based models achieved an accuracy of 98.75% and 96.25%, respectively. This accuracy was considerably higher than that for conventional pattern recognition techniques. The experimental results for the identification of soybean diseases indicated that the proposed model achieved highest efficiency. Keywords: Plant Disease, Deep Learning, Convolutional neural network.