The Research On Level Set Segmentation Method Based On Prior Shape Information

The definition of the segmentation is the process of dividing the image intoseveral uniform and homogeneous regions. The image segmentation is the basis of thecomputer vision and the techniques of the image processing, and provides thetechnical support for the standard of the multimedia data encoding. The situation ofthe study on the image segmentation will directly affect the level of research in theseareas. So, there are more and more attention and concern on the image segmentation.Currently, the geometric active contour model, which is based on the theory of thelevel set, is the more popular image segmentation model. But, when the backgroundof the image is too complicated or the target is obscured by other objects, if thetraditional geometric active contour model is used, then the segmentation results arenot ideal. So, it becomes an import way to introduce the priori information into thegeometric active contour model, and it better improves the segmentation result byintroducing the priori shape. In the algorithm based on Principal Component Analysis,it is to use linear form to reconstruct the samples when dealing with the SignedDistance Functions of the samples, but all of the SDFs of the samples are notrepresented via linear form, which means that most SDFs of the samples are nonlinear,and this will take the drawbacks and shortcomings to the formal segmentation, evenmake the final segmentation results not be satisfactory.In this paper, a novel segmentation model using the priori shape is proposed inorder to solve the insufficient of the segmentation model using the traditionalPrincipal Component Analysis (PCA) algorithm. The major works and innovations inthis paper are:1. The Kernel Principal Component Analysis (KPCA) algorithm is adopted toreduce the dimension of the SDFs of the samples, then solves the correspondingeigenvalues and eigenvectors which are the basis of the proposed shape energy term.This algorithm solves the disadvantages and shortcomings and of the segmentationmodel based on the traditional PCA algorithm to some extent.2. Based on the theory of nuclear feature space, a novel shape energy term isbuilt. The total energy term combines the image energy term which is the improvedLBF segmentation model and the shape energy term which is the novel shape energyterm proposed in this paper.3. A novel pre-segmentation method is proposed in this paper, which meansdoing the pre-segmentation on the image before using the total energy functional tosegment the image. The specific method is firstly to calculate the sample mean of theSDFs of the samples, secondly to do the expansion operation on the solved samplemean to obtain the new sample mean, finally to translate the novel sample mean tomake it closer to the contour of the object. Regard the contour which is obtainedfinally as the initial contour which is the contour of the formal segmentation to make the total energy functional segments the image better and faster.