Watershed segmentation based on gradient reconstruction and region merging

A new method of watershed segmentation is proposed based on gradient reconstruction and region merging. Non-linear transformation and morphological opening-reconstruction and closing - reconstruction are used to modify the gradient image. Furthermore, some small segmentation regions are incorporated into the near bigger regions which is based on criterion of the average gray value of the region, and according to the contrast control and the edge strength criterion. The experimental results show that this method has good robustness and adaptability which can settle the over-segmentation problem effectively.