Abstract

Abstract:Pestdetection techniques is recently increasing in order to grow the productivityand quality of yield. To improve crop productivity disease identification isvery much important, but some diseases cannot be identified at early stage byfarmers and the crop will harm. Hence to avoid the decay of the plants and toimprove quality of the product pest detection is important. The pest detectionmethodology will be developed by using the image processing algorithm like K-means algorithm. It will classify the leaf image in the factors likecolor-features, texture features, ad their combinations to train three modelsbased on support vector machine classifier. The dataset will be trained datasetwhich will consist the thousands of images collected from plat village andfarms. The system will be having the trained dataset to identify the healthyand diseased leaf. It will contain the different images of the diseased leafalong with the healthy leaf. The system will be easier to use as it only takesthe input as image of plant and it is giving result like healthy and diseasedplant. Hence farmers can easily use the system to improve the quality of crop.