Date of Publication : August 2026

Abstract: Cotton is a cornerstone of global commerce and a vital contributor to the agricultural economy. However, the crop is highly susceptible to various pathogens, including Fusarium wilt, bacterial blight, aphid infestations, and the cotton leaf curl virus, all of which pose a significant threat to yield and fiber quality. Manual disease monitoring is labor-intensive and often leads to delayed interventions, resulting in substantial economic losses. To bridge this gap, this research presents an automated, deep learning-based diagnostic system designed for high-precision cotton leaf disease detection. The core framework leverages a Convolutional Neural Network (CNN) for automated feature extraction and classification, distinguishing between healthy tissues and various diseased states. The system processes raw leaf imagery through standardized resizing and noise-reduction pipelines before classification. For practical field deployment, the model is integrated into a responsive Flask-based web application, providing farmers with an accessible interface for instant image uploads and diagnostic feedback. Beyond basic identification, the system incorporates advanced utilities such as disease severity estimation, data-driven visualization charts, and an AI-driven assistant for tailored crop management strategies. Experimental evaluations confirm that the proposed integrated approach offers a robust, scalable, and reliable solution for early-stage disease management, ultimately enhancing crop productivity and sustainable farming practices.

Reference :