Hybrid Deep Learning Model Integrating VGG-19 and ResNeXt50 for Detection of Corn Leaf Diseases to Enhance Global Food Security

Context: Corn diseases are among the significant persistent thorns to global agricultural productivity, directly threatening food security for millions worldwide. As a staple crop integral to nutrition and economic stability, corn plays a critical role in supporting livelihoods and sustaining economies globally. Objective: Traditional approaches to diagnosing corn diseases are often resource-intensive, time-consuming, and inaccessible to many farmers, especially in low-resource settings. This calls for a robust, efficient, and accessible method for the early and accurate identification of corn diseases. Methods: This research handles this problem by designing a hybrid deep learning model, which combines both VGG-19 and ResNeXt50 architectures in an image-classification-based approach to detect the diseases of the corn leaf. Thus, the custom dataset of 15,000 images of six categories- Common Rust, Gray Leaf Spot, Healthy, Northern Leaf Blight, Northern Leaf Spot, and Phaeosphaeria Leaf Spot-covered images were used for training the hybrid model, which was further utilized in a web application for the prediction of diseases in real time. Results and Conclusion: In comparison, hybrid model achieved accuracy of 96%. This is high enough to completely eliminate the possibilities of misdiagnosis even for conditions with overlapping visual features. The web application provides quick and reliable assessment tools for the farmer and agriculture experts to determine timely interventions for optimized crop management strategies. Significance: This study will help to promote sustainable agriculture as it will reduce crop loss and increase yields. The web application supports the United Nations Sustainable Development Goal (SGD), which is hunger removal and improved food security. With technology as the enabler, this solution meets the increased demand for corn while supporting the under-resourced farming communities across the world.