Abstract
Pests that disrupt the roots and leaves of Indian crops are the main cause of their problems. Plant diseases can harm crops and result in financial losses. Improvements are thus required to reduce plant diseases by early detection, which leads to increased productivity. Every year, a large amount of produce is squandered due to the quick invasion of pests and other insects. Due to the wide range of symptoms associated with the diseases, diagnosing infected cotton plants is quite challenging. In this study, we identified the kind of disease and pest in cotton plants using a novel method. Images of diseased leaves will be pre-processed first. After then, images are subjected to edge detection. Advanced fuzzy K-means clustering will be used to segment the edge-detected pictures. After that, the leaf picture's color, correlation, entropy, and texture features—such as energy, contrast, and edges—will be retrieved and compared to a typical cotton leaf image. Ultimately, the precise illness will be shown. Thus, the diagnosis is rather simple. When compared to other methods, the suggested method is very sophisticated and reduces analysis time by 40% when compared to C-means and 52% when compared to K-means clustering. The Matlab IDE is used to write and simulate the code.