Classification of Potato Leaf Disease Using Deep Learning: A Comparative Analysis of custom CNN, ResNet50, and Inception

The potato is one of the most consumed crops globally, grown in diverse climates that bring many unique variations of this essential crop. In many underdeveloped regions, farmers face challenges in identifying and addressing crop diseases due to limited resources and expertise, often resulting in costly losses. Accurate diagnosis is the first step to sustainable crop management, yet most available models only detect a limited number of diseases. To address this gap, our paper presents a custom deep learning model capable of identifying a broad spectrum of issues, including bacterial, fungal, viral, and pest- related diseases. This innovative solution provides farmers with a cost-effective and accessible tool for early disease detection, promoting better crop health and resilience.