Chili Plant Leaf Disease Detection Using SVM and KNN Classification

Abstract

Agriculture is the main segment of Indian economy and driving force for all of the business sector, and nearly about 70% Indian farmers are producing the most needful crops like chili, potato, rice and so on. Being world’s largest spice provider, farmers are planting chilies for exporting worldwide. Also, for every daily food chili is required, hence there is huge need of chili worldwide. Hence it becomes necessary to have automation and live monitoring in chili farming to increase productivity. During production of chili there are several challenges, but disease management is a very critical factor out of it. Currently, there are very few farmers using the technology for disease detection or monitoring, which leads to big losses to production. Hence, we have proposed our new invention in this domain to provide fully automated and easy-to-use chili plant leaf disease detection system using support vector machine (SVM) [2] and K-nearest neighbor (KNN) [1] algorithm. With the help of image classification technique, we have implemented our proposed system and tested it on real-time dataset of various diseases. We have used GLCM [5] feature extraction technique to improve accuracy of detection. Furthermore, we have tested the system based on various case studies and compared the derived outcome with some existing traditional algorithm to prove high efficiency of our proposed system.