Research On Image Segmentation Based On Active Contours Snake Model

Traditional edge detection method can not satisfy the requirement which isn’tincluded in the image itself. The active contour model not only contains the informationof the image itself, but adds prior knowledge from exports. It can accurately locate theboundary of the image.In this paper, we do a deep research on the edge detection theory which is based onthe parametric active contour model and the geometric active contour model. Firstly, areview of image edge detection is presented. It includes the purpose and thedevelopment of image edge detection. And then, the Snake model including its basicprinciple，and Snake model energy function is introduced. To solve the issue of theconcave regional, we further introduce the basic algorithms and numericalimplementation of gradient vector flow. And then, in order to solve the problems of thetraditional Snake model, it should find a suitable way of setting initial contour. Fourthly,we focus on discussing the Snake model based on the level set method, which is ageometric active contour model. In the paper, we use the method of combiningmorphological reconstruction method with level set method in application of human gaitcontour extraction, extraction of number plate, vehicle extraction image field, andachieved good results. The simulation results show that, the algorithm can reduce thenoise in the image and improve the computational efficiency. Finally, we summarize theadvantages and disadvantages of the parametric active contour models and thegeometric active contour model respectively.