Font Size: a A A

Research Of The Organ Segmentation Method Based On The Mouse Model

Posted on:2013-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:Y ChenFull Text:PDF
GTID:2234330395956425Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
It is an important way to use the animal model in the modern biomedical research,it is not only the need of the little animal imaging technology,but also the medicalimage processing and analysis to segment the organs from the medical image.Firstly,in this paper,an interactive segmentation method is proposed which iscombining the fuzzy connectedness theory and Otsu method according to the mousemodel.The dynamic programming is used to find out the fuzzy connection objects,thenbone of the mouse image is extracted by setting threshold using Otsu method.It provesto be effective.In addition,an improved anisotropic smoothing filter is proposed to do somepreprocessing of the image,combining with the histogram analysis method to provide agood initial value for the EM method which is a classical algorithm to solve the mixturedensity according to the medical image segmentation method based on Gaussianmixture distribution,then realizing the decomposition of the mixture density by EMmethod iteratively.Lung of the mouse image is extracted using this method,and itproves to be superior by the experimental results....
Keywords/Search Tags:Medical image segmentation, Fuzzy connectedness theory, Otsu method, EM algorithm, Anisotropic smooth
PDF Full Text Request
Related items