Deep Learning-Based Plant Disease Classification Using VGG-16: Performance Evaluation and Comparative Analysis

Detecting and classifying plant diseases is crucial for effective disease management and preventing significant yield losses in agriculture. Deep learning techniques, particularly CNN, have shown great potential in image-based classification tasks, specifically around plant disease classification in recent years. To decrease computational complexity, the proposed plant disease classification system undergoes pre-processing, which involves resizing and normalization of the input image. During the feature extraction stage, the VGG-16 model is utilized to extract pertinent features from the input image. Ultimately, the extracted features are utilized to categorize the image into one of the probable disease classes by applying a SVM classifier. Upon evaluation on a publicly accessible dataset, the proposed system demonstrated an accuracy of 93%. This study suggests that the VGG-16 architecture can be an effective model for plant disease classification and has the potential to improve the efficiency of disease management in agriculture.