基于模糊颜色提取的彩色图像分割  

Color image segmentation based on fuzzy color extraction

作　　者：张军[1] 李蒙[1] 杨正瓴[1] 

机构地区：[1]天津大学电气与自动化工程学院,天津300072

出　　处：《天津理工大学学报》2014年第6期33-38,共6页Journal of Tianjin University of Technology

基　　金：天津市科技支撑计划重点项目(10ZCKFSF01100);天津市科技型中小企业创新基金(13ZXCXGX40400)

摘　　要：为了对彩色图像进行有效分割,提出了一种基于模糊颜色提取的新型彩色图像分割算法,这种算法可以有效的进行区域分割.首先,提出了一种模糊颜色提取方法,然后介绍了该算法,算法经过迭代可以选择典型的种子颜色来提取颜色分量,最后应用区域生长法可以确保所需区域的完整连通性.区别于现有的图像分割方法,本方法是将相同的像素分成几个模糊集,分割成功率提高10%左右.实验结果表明该算法可以合理分割出所需区域,避免冗余,并且具有较好的噪声抑制力和鲁棒性.In this paper, we proposed a new color image segmentation algorithm based on fuzzy color extractor to segment out meaningful region. Firstly put forward a method for extracting fuzzy color, and then introduces the algorithm, the algorithm through iterative can choose typical seed color used to extract the color components. Finally the application of region growing method can ensure the desired area be connective. Different from other segmentation method, this method is put the same pixel divided into several fuzzy sets, the success rate of segmentation improved about 10%. The experimental results show that the algorithm can reasonably required region segmentation, avoid redundancy, also has a stronger anti-nose ability and good robustness.

关 键 词：彩色图像分割 模糊颜色提取 区域生长法 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]