A novel watershed image segmentation technique using graceful labeling

In this paper, we propose a novel watershed image segmentation technique based on absolute difference graph labeling. The main objective is to partition the image into distinct regions that are visually different, homogeneous and meaningful with respect to some characteristics or computed property, that is, such as grey level, texture or color to enable easy image analysis. A graph representing an image is derived from the similarity between the pixels and partitioned by a computationally efficient absolute difference graph method, which identifies representative vertices for each absolute difference and then expands them to obtain complete absolute difference graph labeling. As a result, the image is portioned into a set of disjoint labels (regions).