Font Size: a A A

Medical Image Segmentation Method Based On Level Set

Posted on:2016-07-12Degree:MasterType:Thesis
Country:ChinaCandidate:S H LvFull Text:PDF
GTID:2308330482474769Subject:Surveying the science and technology
Abstract/Summary:PDF Full Text Request
In recent years, the rapid development of information technology and the increasing levels of computer applications prompt the advances in medical imaging technology.The role of medical images has changed from the traditional visual observation of anatomy to the use of computer graphics and image technology, human anatomy for the region of interest, such as lesions automatically accurate positioning, segmentation, feature extraction and quantitative analysis and other treatment,which has made computer medical image processing technology being a hot research topic at home and abroad in recent years.And the medical image segmentation is the key technology of medical image processing.Compared to other images,the complexity and diversity of medical image make it difficult to obtain good segmentation results of the traditional segmentation method.Active contour model based on level set which is closer to the human visual mechanism, and do not track curve topology change in the evolution of the process, and will not let the phenomenon to occur intermittent border not only use the underlying image information, but also a combination of a priori knowledge of high-level. This makes the level set method has been widely applied in the field of medical image.In this paper, I make a deep study of medical image segmentation method based on level set, including the following aspects:The first one, a representative of the level set segmentation method: GAC method, Li model approach, improved Li model approach,C-V method and V-C multiphase level set method,and applying them to artificial image and medical image segmentation. The second one, introducing improved distance penalty term into the C-V model can avoid re-initialize the level set function and enhance the speed of evolution. This paper proposed general model and evolution equation of arbitrary level set in multiphase level set model which Introduced a distance penalty term. The third one, aiming at the advantages and shortcomings of level set segmentation based on edge detection and region,I proposed the combined method which combine these two method, and we obtained the better segmentation by experiments.
Keywords/Search Tags:Medical image segmentation, Level set methods, Active contour model, Chan-Vese model, Distance penalty term
PDF Full Text Request
Related items