Implementation of a CNN-Based Model for Soybean Leaf Diseases

Soybean ranks among the most vital crops grown in India, especially in Madhya Pradesh, where it significantly contributes to the agricultural economy and supports nutritional security. Nonetheless, the yield of soybean crops is greatly affected by several leaf diseases, including bacterial blight, downy mildew, soybean rust, southern blight, and powdery mildew. Timely and precise detection of these diseases is crucial to reduce crop damage. Conventional disease identification techniques are often slow, require considerable manual effort, and are susceptible to human error. To address this issue, this study proposes the implementation of a Convolutional Neural Network (CNN)-based deep learning model to detect and classify common soybean leaf diseases using image data. A total of 5,917 images were utilized in the dataset, divided into five disease categories and one healthy category. The dataset was preprocessed and augmented to enhance model performance and divided into training (70%), validation (10%), and testing (20%) sets. The CNN model was trained over 20, 40, and 50 epochs to assess its performance across varying training durations. It demonstrated high classification accuracy, highlighting its effectiveness as a dependable method for the early detection of soybean leaf diseases.