Performance evaluation of color image segmentation using K means clustering and watershed technique

Image segmentation is a key technology in image processing which partition an image into its constituent regions. Watershed and k means segmentation techniques are practical approaches for color image segmentation. This paper discusses quantitative evaluation measures for color image segmentation based on these techniques. Color image segmentation can be viewed as an extension of gray level image segmentation. Quantitative measures like discrete entropy, root mean square error, visible color difference are proposed for color images.