Research On Image Segmentation Of Active Contour Model Based On Level Set Method

With the rapid development of electronic science of the twentieth century, more and more image segmentation applied to the actual life. Image segmentation had been the hotspot and difficulties of image processing techniques.Firstly, this paper mainly introduced the background of image segmentation and its significance. In addition, the existing image segmentation technology based on partial differential equations were reviewed, such as GAC model, CV model and ICP model. Their advantages and disadvantages were compared.Secondly, SPF method was introduced and analyzed. According to the characteristics of SPF function and the kernel function, the new SPF function with local intensity clustering properties was proposed. Using the penalizing term, the new model ensured the adaptability of level set function. Compared to the SPF method, the model was only sensitive to the target area and efficiency on intensity inhomogeneities. Besides, the function form was simple. Compared to the ICP model, it could deal with weak boundary better. In addition, the method initialize simply and iterate rapidly.Then, this paper combined IAC model and ICP model after introducing IAC model. And this paper proposed an improved method based on the edge information and regional information. Using the key role of penalizing term, it could ensure the adaptability of the level set function. And the method initialized simply and iterated rapidly. Compared to the IAC model, the model could efficiency on intensity inhomogeneities. Compared to the ICP model, it dealt with weak boundary better.Finally, it summarized the research work briefly, and pointed out the direction of further work.