Cotton disease detection based on deep learning techniques

Cotton is one of India's most famous cash crops. Cotton production is decreasing last year as a result of the disease's invasion. Generally, these plant diseases are caused by insects or pests, which reduce the production on a large scale if not restricted promptly; cotton plant leaf diseases must be identified early and accurately, as they can harm yield. Identifying and managing the disease has become a significant challenge due to the rapid growth of various diseases and a lack of adequate knowledge among farmers. The leaves have a similar texture and appearance, which helps to identify the disease type. As a result, deep learning with computer vision offers a solution. A deep learning model is trained with healthy and infected images is proposed in this paper. This model achieves its goal by classifying leaf images into different categories like healthy or diseased based on the infected patterns. The proposed method's performance is evaluated on the Cotton Disease Dataset Kaggle and the classification accuracy is achieved 97.13%, which is better than the existing state-of-the-art methods.