Evaluation of the Performance of Different Deep Learning Models for Classifying Corn Leaf Infections

Corn, a globally significant crop, serves as a cornerstone in the production of both food and feed. However, the susceptibility of corn plant leaves to infections caused by various pathogens poses a threat to crop yield and quality. Traditional manual inspection methods are laborious and prone to errors, necessitating advanced approaches for timely and accurate detection. This research addresses the challenge by leveraging deep learning techniques, specifically convolutional neural networks (CNNs), including ResNet50, VGG16, Inception V4, and AlexNet, and the research explores transfer learning techniques and systematically fine-tunes model configurations. The proposed methodology involves training and optimizing these models on a publicly available dataset of labeled corn leaf images, employing metrics such as accuracy, precision, recall, and F1 score for evaluation. Results reveal ResNet50 as the top-performing model, achieving an exceptional accuracy of 99.76 %, showcasing its prowess in accurately classifying infections. This research attempts to enhance sustainable agriculture by improving crop yield, quality, and ultimately, food security.