灰关联空间中基于最大熵阈值的医学彩色图像分割算法  

A medical color image segmentation algorithm based on maximum entropy threshold in the grey space

作　　者：齐敏[1] 邓林[1] 弋丹[1] 宣杭[1] 郝重阳[1] 

机构地区：[1]西北工业大学生物医学工程研究所西北工业大学虚拟现实研究中心,陕西西安710072

出　　处：《中国科学技术大学学报》2007年第12期1543-1545,共3页JUSTC

基　　金：教育部博士学科点专项基金(20040699015)资助

摘　　要：将灰关联理论引入彩色图像分割领域,提出一种在灰关联空间中面向医学彩色图像的最大熵阈值分割算法.通过对分辨系数和色彩影响因子的选取,避免了传统彩色图像分割中三维色彩信息的丢失现象,并使算法具备了一定的抗噪性能,分割的精度和实时性也获得了较大提高.实验结果证明了算法的有效性.The theory of grey relation analysis was used to segment color images in a creative way. An algorithm for medical color image segmentation based on maximum entropy threshold and the concept of color grey relational degree images was proposed. The method prevented the problem of color information loss existent in the previous color image segmentation algorithms and some anti-noise capability was obtained by choosing the resolution coefficient and color effect gene. At the same time, the accuracy was increased and runtime was shortefied. The experimental results indicate the effectiveness of the algorithm.

关 键 词：彩色图像分割 灰关联分析 色彩灰关联度图像 最大熵阈值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]