Cotton Leaf Disease Detection Using Transfer Learning

Abstract: Cotton is a critical crop in the agricultural sector, contributing significantly to the global economy. However, its productivity is frequently hindered by diseases that affect the leaves. Early detection of these diseases is essential to minimize losses and ensure sustainability. This paper presents a solution using transfer learning to detect and classify cotton leaf diseases with high accuracy. The proposed system employs pre-trained models like Mobile Net, VGG16, and ResNet152V2, fine-tuned to a dataset of cotton leaf images. Results demonstrate an accuracy of up to 99.32% using Mobile Net, highlighting the system's effectiveness and feasibility for real-world applications.