基于二次聚类的主动脉弓分割方法  

Segmentation Method of the Aortic Arch Based on Quadratic Clustering

作　　者：陈中中[1] 杨亚茹[1] 张建飞[1] 王倩倩[1] 朱惠玉[2] CHEN Zhongzhong;YANG Yaru;ZHANG Jianfei;WANG Qianqian;ZHU Huiyu(Mechanical Engineering,Zhengzhou University,Zhengzhou 450001,China;Productivity Promotion Center,Zhengzhou 450001,China)

机构地区：[1]郑州大学机械工程学院,河南郑州450001 [2]河南省生产力促进中心,河南郑州450001

出　　处：《郑州大学学报（工学版）》2018年第3期40-44,共5页Journal of Zhengzhou University（Engineering Science）

基　　金：国家科技支撑计划项目(2014BAI11BO8-2-02)

摘　　要：提出一种基于均值漂移和层次聚类的二次聚类图像分割算法(MSHC),在CIE(LUV)颜色空间,首先运用均值漂移算法实现图像的平滑及初步聚类,然后将结果中每个区域的均值作为第二次聚类的初始值,对图像进行层次聚类处理至满足预定效果为止.最后,将图像中主动脉所在类的均值作为区域生长的种子点,完成目标提取.该MSHC算法分割效果良好,易于重建主动脉弓三维模型,且模型立体感强,可清晰展现其空间三维结构.A twice clustering method（ MSHC） was proposed based on meanshift and hierarchical. Firstly,in the CIE（ LUV） color space,meanshift method is used for the first clustering to realize image smoothing and initial clustering. Then,the mean of each region was used as the initial value for the second clustering. And hierarchy clustering was used for the second clustering,which would be stopped until the clustering effect reached the expected effect. Finally,the mean of the aorta region was used as the seed point in the region growing,and the image segmentation was finished. The MSHC method could improve the efficiency and effect of image segmentation,and the aortic arch model had a strong visualizing and realistic sense,which could clearly show the three-dimensional structure.

关 键 词：均值漂移 层次聚类 二次聚类 主动脉弓 三维重建 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]