A Deep Learning Approach to Detect and Classify Wheat Leaf Spot Using Faster R-CNN and Support Vector Machine

The occurrence of fungal disease in wheat crops, known as wheat leaf spot disease, is a frequent issue that can result in substantial decreases in crop yield. Early detection and accurate classification of the disease severity are essential for effective disease management. In the proposed work, a hybrid model using Faster region based convolutional neural networks (R-CNN) has been developed and support vector machine (SVM) for the detection and classification of wheat leaf spot disease. We collected a dataset of 10,000 images and performed binary classification to distinguish between healthy and diseased wheat leaves, achieving an accuracy of 96.63%. We also conducted multi-classification to categorize the disease severity of the wheat leaves into five different levels, achieving an accuracy of 96.33%. The results of our study demonstrate the effectiveness of the hybrid model in accurately detecting and classifying wheat leaf spot disease. The high accuracy achieved in both binary and multi-class classification indicates that the model is robust and reliable. The agricultural industry can benefit significantly from our findings since the precise classification and timely identification of plant diseases can prevent substantial crop damage. We believe that a hybrid Faster R-CNN and SVM model offers a promising method for identifying and categorizing wheat leaf spot disease. The findings open up new avenues for agricultural study and demonstrate the potential of machine and deep learning (DL) techniques.