| 改进的基于区域合并的纹理图像分割方法 |
| |
| 引用本文： | 郑庆庆,桑农,高常鑫,刘乐元.改进的基于区域合并的纹理图像分割方法[J].华中科技大学学报(自然科学版),2011(5):109-112. |
| |
| 作者姓名： | 郑庆庆 桑农 高常鑫 刘乐元 |
| |
| 作者单位： | 华中科技大学图像识别与人工智能研究所; |
| |
| 基金项目： | 国家高技术研究发展计划资助项目(2009AA12Z109); 教育部新世纪优秀人才支持计划资助项目(NNCET-05-0641) |
| |
| 摘 要： | 针对自然纹理图像的特点,提出了一种改进的基于区域合并的纹理图像分割方法.首先选择符合人类视觉对颜色的感知区分能力的L*a*b*颜色特征;然后提取图像的Gabor能量作为纹理特征;接着由颜色相似度和纹理相似度的概率加权平均获得2个相邻区域的相似度;最后利用基于最大相似度的区域合并算法交互式地完成图像分割任务.实验结果表明:该方法比仅使用红绿蓝(RGB)颜色特征的相似度测量获得了更加精确的分割效果,并且在相同的初始过分割以及人工交互条件下,优于Lazy Snapping.
|
| 关 键 词： | 纹理图像分割 颜色空间 Gabor能量 最大相似度 区域合并 人工交互 |
Improved textured image segmentation method by region merging |
| |
| Zheng Qingqing Sang Nong Gao Changxin Liu Leyuan.Improved textured image segmentation method by region merging[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2011(5):109-112. |
| |
| Authors: | Zheng Qingqing Sang Nong Gao Changxin Liu Leyuan |
| |
| Affiliation: | Zheng Qingqing Sang Nong Gao Changxin Liu Leyuan(Institute of Pattern Recognition and Artificial Intelligence,Huazh ongUniversity of Science and Technology,Wuhan 430074,China) |
| |
| Abstract: | According to the characteristics of natural textured im ages,a novel image segmentation method based on region merging was presented.T he color feature of L*a*b* color space which accords with the color resolu tion of human visual perception was chosen.Then Gabor energy was extracted as a texture feature,and the similarity of two regions was measured by using the pr obability-weighted average of L*a*b* color similarity and texture similar ity.Finally,the maximal similarity based the region merging algorithm... |
| |
| Keywords: | texture image segmentation color space Gabor energy maximal similarity region merging user interaction |
| 本文献已被 CNKI 等数据库收录！ |
