Contour Extraction of Device Wall Image Based on GVF Snake

Active contour model is mainly used in computer vision and image processing,especially in contour extraction.Presently it is a popular image segmentation algorithm,and its advantage is that it always can get smooth and closed border whether the quality of the image is good or not.The device wall image is segmented with Gradient Vector Flow(GVF) algorithm.GVF algorithm can overcome some defects of the traditional active contour model,such as it can′t detect sunken part of image,has a low speed of convergence and its initial contour must be close to contour of object.Experiment results show the algorithm can not only get better effect of segmentation to the device wall image,but also improve the speed of contour extraction.Therefor it provides a good environment for calculating the geometric parameters of the damaging parts of device wall image.