SIFT Descriptor Based Early Diognosis of Leaf Diseases

摘要

This paper proposes a robust leaf diseases classification system based on Scale-Invariant Feature Transform (SIFT) descriptor. The system will segment the diseased region in the input image by using HSV based image from the diseased part of the image are grouped by the K-means method to obtain the basic feature or visual word. The obtained feature list is converted into a Bag-of-visual features, which are then classified by using K-means clustering to categorize a particular disease in the leaf image into their appropriate group.