The Application of Watershed Algorithm in the Initial Contour Selection of Mumford-Shah Model

The image segmentation based on active contour model is currently a hot-spot issue, among which Mumford-Shah model is an important method. To solve the problem of unbearable computing time, this paper intends to combine Mumford-Shah model and watershed algorithm in the image segmentation. Specifically, the image will be segmented firstly through Watershed algorithm and then the border produced will be the initial contour of Mumford-Shah model. Since watershed algorithm has the advantage of segmenting the target object from the complex background, the contour produced thereupon is very close to the border of the target object. After that, Mumford-Shah algorithm is used to segment the image with the contour produced thereupon as the initial contour. In this way, the iterative time is greatly reduced and a good result can be expected