Advancing Precision Farming with AlexNet: A Scalable Deep Learning Model for Maize Leaf Disease Identification

Maize leaves disease poses a severe threat to worldwide agricultural productivity, particularly in mixed illumination field conditions, composite backgrounds, and mixed leaf appearances. While tremendous progress has been made in plant disease diagnosis with deep learning-based approaches, there are still major hurdles in model robustness and providing consistent performance on a broad range of agricultural conditions. In this paper, we present an improved AlexNet-based convolutional neural network architecture specifically tailored for accurate and efficient classification of four unique maize leaf conditions: Corn Blight, Common Rust, Gray Leaf Spot, and Healthy. The model was thoroughly trained and tested on a well-curated maize leaf dataset with training accuracy of 97.08% and testing accuracy of 94.48%, and precision rates of 93.78% and 88.95%, respectively. False positive rate was limited to 3.85% during testing, and inference time was reduced to 2.2 ms/image, thus establishing the possibility of real-time applications. Further, the framework converged stably within 30 epochs, establishing its computational efficiency. This work bridges major gaps of existing models by improving resistance to environmental variability and reducing misclassification. The proposed method provides a strong benchmark for deep learning-based plant disease diagnostics and establishes the efficacy of optimized CNN architectures like AlexNet in facilitating scalable and intelligent solutions for precision agriculture.