The potato, is one of the most important agricultural crops, especially in impoverished countries. Fungal infections pose a substantial global risk to potato crops. Despite facing challenges related to computation time and accuracy, have explored the realm of machine learning for early detection. In this study, the aim is to introduce a modified convolutional neural network (CNN) to enhance precision and reduce the computational burden in identifying fungal infections in potatoes. The model’s performance undergoes a thorough comparison with existing techniques for potato categorization. With the overarching goal of supporting the resilience and success of global potato farming, the research strives to simplify and improve the detection and control of fungal infections.