K-Nearest Neighbor Algorithm based Classification of Cauliflower Pest for Pest Identification in Vegetation

Abstract

            Agriculture is meant for producing sufficient food for the growing population. But today the agriculture is widely affected by change in climatic conditions, Pest and Diseases. Pest problem is the major problem in agriculture. Pesticides are utilized to control the pest in the fields. These pesticides are dangerous which not only affects the harmful pests but also affects the useful pest. It also makes the farm infertile. Therefore, a technique which inspect and eradicate the pest without harming the environment has been identified. In this paper the pest infecting the cauliflower are identified by using image processing technique. More than 100 images of pest infecting the cauliflower are gathered from various farms of cauliflower in real time for the analysis. Enormous factual highlights are extricated from the pictures which will improve the precision in recognition of pest. The effective testing and training has been done. The K Nearest Neighbor classifier is used in this work for pest classification. This technique detects the pest in its initial stage. Hence the control action has been taken in time without damaging the crops. The MATLAB is the simulation tool which is being used in this paper for validation of results.