Reseach On Image Segmentation Algorithm Based On Semi-Supervised Clustering

Image segmentation is the most important issue in image analysis and pattern recognition, and it largely determines the final image quality of the analysis and discriminant analysis results.Semi-supervised clustering algorithm is a research focus in the international field of machine learning and data mining, it attracted many scholars to study in this field, and have achieved some results.In this paper, the image segmentation algorithm and semi-supervised clustering algorithm are studied, and two new image segmentation algorithm based on semi-supervised clustering are proposed. The segmentation results are verified by experiments.This paper, to a certain extent, riches image segmentation algorithm for the research contens.It provises a new train of thought for the image segmentation problem solving, and has a certain scientific value and potential application.In the process of semi-supervised clustering,the sample data with labels or constrained information are auxiliary for the processing of clustering. So, the most important problems in semi-supervise clustering is how to use the sample datas with labels or constrained information effectively to guide the clustering process for getting the more better result clusters. Image segmentation algorithm based on semi-supervised clustering algorithms improves the traditional image segmentation algorithm based on clustering, integrated limited artificial supervision information, which is artificial in-band image segmentation,click a few points, these points can be as much as possible on behalf of all regions of the information necessary to split out, these points can identify the relationship between the various regions as the monitoring data points in the semi-supervised learning with label information, with these small amount of supervision and data points with the label information on a large number of tag information to be split data points split, thereby increasing the accuracy of the algorithm.From the structure, this paper introduces the basic theory and common method for image segmentation, including traditional graphic segmentation methods and image segmentation method combined with a specific theory, the proposed two semi-supervised clustering-based image segmentation algorithm is combined with the image of a particular theory segmentation method based on clustering method based on improved. The third chapter in this paper introduces the related theory about clustering firstly, the basic concepts in clustering are explained. Meanwhile, common types of clustering cluster and common clustering algorithm are also introduced. In subsequent content, this paper focuses on the revelant knowledge about semi-supervised clustering, some kinds of measure methods for clustering, including their principles.In the fourth chapter, this paper proposes the core content, two new image segmentation algorithm based on semi-supervised clustering. In this chapter, the paper shows the main component of the algorithm from three directions in details, include the meanings, process, varification experiments of it. Those experiments show that, these two algorithms has a certain promotion both on clustering accuracy and compactedness.At last summarized to the full paper, and indicated the next work. Then proposed the own views about the further development of the image segmentation algorithm based on semi-supervised clustering.