Hand Segmentation using Modified K-Means Clustering with Depth Information and Adaptive Thresholding by Histogram Analysis

In this paper we present a methodology for segmentation of hand images using modified K-means clustering with depth information of an image and adaptive thresholding by histogram analysis. We extract the hand area by using K -means clustering to divide image into different clusters based upon its intensity value. Thus we can say suggested methodology will give desired results for segmentation of hand images in different conditions like hand color, scale, rotation, pose, lightning conditions and colored background.