Harnessing Deep Learning With AlexNet for Tomato Leaf Disease Detection in the Indian Himalayan Terrain

Agriculture is essential for living in the Indian Himalayan region (IHR), as it functions as the main occupation and source of income. Therefore, monitoring crops and detecting diseases at early stages becomes essential. Our research leverages deep learning (DL), a dynamically growing technology proficient in handling large datasets. In this work, convolutional neural networks (CNNs) are used to solve real‐world problems using the computer vision technique. This paper proposes a modified AlexNet architecture for detecting diseases through images at an early stage and to increase the production rate by cultivating good quality crops. This model combines steps such as tomato leaf image data collection, image data preprocessing, and classification and detection of disease. In this work, a total of 5500 tomato leaf images have been taken from Kaggle, and an additional 1650 photographs have been collected from a variety of field sites, thereby creating a comprehensive dataset that will facilitate the development and assessment of a robust model. In the classification and detection of tomato plant leaf diseases, our proposed model achieves an accuracy of 94.80%. Using DL technology, we hope to give farmers an accurate way to monitor and protect their crops, ultimately enhancing agricultural quality and output.