Research On Image Segmentation Algorithm Based On Superprixels And Region Merging

Image segmentation is an important research field in Digital Image Processing,it is the key step of image processing link to image analysis.The accuracy of subsequent image analysis and image comprehension will be directly affected by the results of Image segmentation.Therefore,the in-depth study of image segmentation is of great significance.Superpixel segmentation methods are image pre-processing technologies that have rapidly developed in recent years.These methods can segment an image into a certain amount of sub-regions which are consisting of adjacent pixels having similar feature.Compared with traditional pixel-based processing methods,superpixels have the advantage of better extraction of local features and better expression of image structure information.Furthermore,superpixels can reduce the size of the processing object while speeding up the computational efficiency of subsequent processing.Given these significant advantages,superpixels have been widely used in many fields of computer vision,and have more and more become a research focus in Digital image processing.Through studying image segmentation algorithms based on superpixels and region merging,this thesis proposes a new image segmentation algorithm called MSBRM(Mean Shift based Bayesian Region Merging)based on Mean Shift algorithm and Nonparametric Bayesian Clustering Model.Firstly,over-segment an image to superpixels using Mean Shift algorithm.Then,a merging criterion based on Nonparametric Bayesian Clustering Model and the spatial information of superprixels is proposed,which is used to form the final segment result by merging the superprixels.Experimental results show that MS-BRM algorithm solves the problem of superpixels over-segmentation,and the final segment result of the images reserves many boundary information,which highly matches human vision.In addition,this thesis also applies the MS-BRM algorithm to the field of human image segmentation,proposes an new human image segmentation algorithm based on face detection and the MS-BRM algorithm.Firstly,use face detection algorithm to recognize human faces and gets facial contours feature points of a human image.Then,use these feature points to establishes seed point estimation model and gets the object and background seed points.At last,use these object and background seed points to merge the human body region and the background region of the initial segmentation of human images made by the MS-BRM algorithm.This algorithm is an automatic human image segmentation algorithm,the segmentation process does not require user interaction and parameter selection.It can get high quality human body region from the background of a human image.