Applications Of Medical Image Segmentation Based On Multi-pass Level Set Method

It can be said that without the medical image segmentation methods, then wewill not be able to analysis and process medical images automatically by computer.Therefore, the medical image segmentation techniques played a very important role,it has been became one of the most hottest research field on medical image.This paper made a deep research on the key technology of medical imagesegmentation, in which the weak boundary image information in the model, thephenomenon of weak boundary in image and regional division for the medicalimage has been included. At the same time, we get more effective information andreduce the computational time of the models in medical image segmentation. In thispaper, two new models has been proposed for solving these problems:The M-F model is for solving the problem of medical image with weak edgeand heterogeneous regions or the problem of medical image with active contoursmodel with multi-channel needs preprocessing (noise reduction) and other issues,by local entropy to enhancement processing capacity of the weak boundary andheterogeneous region of the proposed model.The M-L model is proposed based on the LCV model. A new boundaryfunction is adopted to control the rate of evolution of the level set function in theweak region on the boundary in the model. Also a new penalty function is adoptedto eliminate the periodic phenomenon in initialization and a regional limit functionis introduced to realize multilayer image segmentation.Finally, the proposed algorithm has been simulated by Matlab, and the imageobtained by the visual effects and the experimental data with the traditionalsegmentation segmentation algorithms were compared. In this paper, the algorithmfor image segmentation model can better deal with weak edge section, the model is robust to noise and uneven grayscale image segmentation capability, initialize thelevel set function does not affect the segmentation results, and high efficiency.