Fortifying Tomato Agriculture: Optimized Deep Learning for Enhanced Disease Detection and Crop Health Management

In Indian Cuisine, tomatoes play a substantial role on an everyday platter. India is one of the largest producers of tomatoes in the world, nevertheless, the stature and volume of tomato yield befall accompanying numerous recent diseases. Tomatoes may help fight various diseases in human health, but it is appropriate to save tomatoes from getting affected by various diseases. Although current CNN models are efficient in identifying diseases, it is difficult to accurately detect diseases with little computational complexity. Based on the voting strategy and their various outcomes an ensemble-based disease detection technique with optimized hybrid deep learning is designed to enrich the detection accuracy. A deep learning-based CNN model, including EfficientNet, ShuffleNet, and ResNet-50 categorizes the diseases using the following steps: acquiring plant leaf photos, preprocessing the images, segmenting the images, extracting features, and categorizing the various diseases. The effectiveness of this study is assessed using various evaluation metrics and among the CNN models evaluated, EfficientNet achieved the highest accuracy at 98.74%, followed by ResNet-50 at 97.97%, and ShuffleNet at 96.88%. Enabling farmers to achieve better crop yields and enhance food production, this research assists in early disease detection in plants with minimal computational complexity.