Description of Invention
Early detection of plant diseases is essential for improving crop yield and ensuring sustainable agriculture. Studies have shown that traditional disease diagnosis based on visual inspection is often time-consuming and prone to errors, especially under field conditions. This study applies a convolutional neural network (CNN) to classify tomato leaf diseases using a dataset of 8000 images across 10 categories, including diseases like Tomato Mosaic Virus, Bacterial Spot, and Late Blight, along with healthy leaves. The model consists of three convolutional layers with max-pooling, followed by a dense layer and a softmax classifier. Using data augmentation and rescaling techniques, the model achieved 95.74% training accuracy and 89.50% validation accuracy. These results demonstrate the effectiveness of deep learning in distinguishing visually similar plant diseases. This research highlights the potential of CNN-based disease classification as a valuable tool for precision agriculture, supporting farmers with timely and accurate diagnosis. Future work will explore model optimization for mobile deployment, enabling real-time disease detection in the field.