An Improved Watershed Algorithm for Color Image Segmentation

An adaptive marker extraction-based watershed algorithm is proposed to overcome the over-segmentation problem. By combining local minima depth and water basin scale information, markers are adaptively extracted for local minima, and the threshold for marker-extraction is automatically calculated according to the statistics of local extreme points in the gradient map. These markers are imposed on the original gradient map as its local minima. The watershed algorithm is applied on the modified gradient map to segment the image. Simulation results show that the proposed method can efficiently reduce over-segmentation with scarcely computational complexity increase. It has better anti-noise performance and edge-location capability as well.