详细信息
基于可视色差与区域生长的彩色图像分割
A Color Image Segmentation method based on visible color difference and region growing
文献类型:期刊文献
中文题名:基于可视色差与区域生长的彩色图像分割
英文题名:A Color Image Segmentation method based on visible color difference and region growing
作者:于丽杰[1,2];李德胜[2];王跃宗[2]
第一作者:于丽杰
机构:[1]北京联合大学自动化学院;[2]北京工业大学机电技术研究所
第一机构:北京联合大学城市轨道交通与物流学院
年份:2010
期号:10
起止页码:12-14
中文期刊名:微计算机信息
外文期刊名:Microcomputer Information
语种:中文
中文关键词:可视色差;区域生长;子块;彩色分割
外文关键词:visible color-difference region growing sub-blocks color image segmentation
摘要:提出了一种基于可视色差和分块区域生长的彩色图像分割算法,首先将待处理图像划分成若干大小相同的子块,在CIEL*a*b*颜色空间中计算每个子块的色彩均值和方差值,根据方差值对子块分类,利用可视色差对颜色均匀子块分类、合并,获得初始分割区域,利用四叉树算法对纹理子块分裂合并,然后从初始分割区域开始生长、合并,并删除小颜色块。实验结果证明了算法的有效性,分割结果符合人的主观感知。
The article brought forward a color image segmentation algorithm based on visible color difference and sub-block region growing techniques.Firstly,the original image is first divide into sub-blocks which are not overlapped.In CIELab space,calculate the mean and variance of every sub-black,classify the sub-block with the variance of it,it achieve initial region by classifying and mergeing the color uniform sub-blocks with visible color-difference,split and merge the texture sub-block with quadtree,final grow,and merge from initial region,and delete mini color block.The experimental segmentation results hold favorable consistency in terms of human perception,and confirm effectiveness of the algorithm.
参考文献:
正在载入数据...