Comparative Analysis of Convolutional Neural Network Models for Early Detection of Corn Leaf Diseases: Inception vs. MobileNet

Corn is one of the major crops in India. In agriculture, early diagnosis of corn leaf disease is essential for increasing the yield of crops. One of the most widely grown grains worldwide is maize. Several leaf diseases, including corn common rust, corn grey leaf spot, and corn leaf blight, are prevalent and can cause significant damage to corn harvests. In their early stages, the symptoms of many leaf diseases are hardly distinguishable. This proposed work is a comparative analysis of two prominent convolutional neural network models, Inception version 3 and MobileNet version 2, for the corn leaf disease detection dataset. The research seeks to illuminate their respective performance and computational characteristics by employing the same dataset and consistent settings, including hyper parameters, data augmentation methods, and simulation parameters. MobileNet emerges as a frontrunner in training accuracy, boasting an impressive 95 percent, compared to Inception's 89 percent0. This study substantially contributes to agricultural disease detection by offering empirical insights and a foundation for enhanced model development and real-world applications.