A Novel Geometrical Features for Identification of Diseases in Paddy Leaves

Abstract

Abstract— Novel geometrical features have been proposed for identification of diseases in paddy leaves. Identification of fungal diseases like blast and brown spot diseases are done by using geometrical features and derived features of geometrical features. The Gamma technique is used for enhancement of images. The enhanced RGB images are converted to HSV color space. The Hue component of HSV is considered for segmentation of diseased portion. The kNN classifier is used for classification which has given accuracy rate of 81.18%.    

Keywords- Brown Spot Disease; Blast Disease; Paddy leaves; kNN Classifier; Segmentation, Geometriccal features;