Improvement Of Level Set Algorithm And Its Application In Medical Images

Due to the importance of image segmentation in practical applications,scholars at home and abroad have conducted in-depth research.So far,scholars have achieved many research results in this field,but still can not find a universal segmentation algorithm to adapt to all types of images.However,in the field of image segmentation(especially medical image segmentation),the level set algorithm has its unique advantages.The result of this algorithm and its performance all keep ahead,so this method is also the most promising to become a general segmentation method.Because the formation of medical images is affected by noise and tissue motion,the gray scale is not uniform and the boundary is not clear,so it is difficult to segment medical images.Until now,the level set method theory has been basically solid,but there are still some problems that make this method not perfect.Therefore,this paper improves Chan-Vese(C-V)level set algorithm in image segmentation.The main tasks as follows:(1)The results obtained by scholars at home and abroad in the level set algorithm in recent years are studied,and the scope and defects of the application are analyzed.(2)The level set algorithm is studied and implemented,applied to image segmentation,and the segmentation results are analyzed.(3)The Mumford-Shah algorithm is studied,and the C-V level set algorithm is described and implemented on the basis of this.It is applied to image segmentation and the experimental results are analyzed to obtain the advantages and disadvantages of this model.(4)A C-V level set algorithm based on Canny operator is proposed.Experiments show that the proposed algorithm not only makes the segmentation result more accurate,but also reduces the interference of noise on segmentation.For some images with weak edge information,the region information can be fully utilized to achieve accurate segmentation.