Research On Multi-region Image Segmentation

A image can be segmented according to some characteristic into some regions that have some meanings and will not be superposed by a method of image segmentation. Image segmentation is a pivotal technique from image processing to image analysis. Quality of image segmentation directly affect the following result of image processing, therefore the image segmentation technique is regarded as the bottleneck of image processing and is very important.In this paper multi-region segmentation of a gray scale image is mainly studied. First a image is preprocessing, linear or nonlinear filtering of spatial domain is applied to the image to remove the noise and the algorithm of edge detection is applied to the image to enhance the edges of image. They are all preparations for the image segmentation. There are three different methods applying to image segmentation, which are respectively based on regions, the choice of threshold and the employment of special theory.In the methods of image segmentation based on region, a region growing algorithm, splitting and merging algorithm, relaxation iteration algorithm, K-means clustering algorithm and comparability segmentation algorithm are adopted,_some parameters is improved through experiment.In method of image segmentation based on the choice of threshold, traditional iteration threshold algorithm and method of single threshold based on gradient, concavity analysis, moment and entropy are improved in the paper. All confines of gray scale are segmented into some parts, and multiple threshold is derived through in each part histogram of gradient, concavity analysis of histogram, moment and entropy of histogram. The multiple threshold image segmentation is achieved.In this paper the wavelet analysis technology and the data fusion technology is applied to image segmentation. In method of image segmentation based on wavelet analysis, multiple threshold is obtained from the histogram of gray scaleby multi-resolution analysis. The threshold is applied to image segmentation. In method of image segmentation based on data fusion, the image is respectively processed by edge detection algorithm and segmentation based on region, and these two images are syncretized into a new one after image rectification.Experimental results show that the image is distinctly segmented into multi-regions that have different character. Especially after applying method based of data fusion, the image is not only distinctly segmented into multi-regions, but the edge of the image would be legible enough.