| 图象过渡区提取与分割算法评价 |
| |
| 引用本文： | 闫成新,桑农,张天序. 图象过渡区提取与分割算法评价[J]. 计算机工程与应用, 2005, 41(8): 60-63 |
| |
| 作者姓名： | 闫成新 桑农 张天序 |
| |
| 作者单位： | 石油大学(华东)机电工程学院,山东,东营,257061;华中科技大学图像识别与人工智能研究所,武汉,430074 |
| |
| 基金项目： | 国家自然科学基金重点资助项目(编号:60135020)，石油大学校博士基金资助项目(编号:Y040313) |
| |
| 摘 要： | 对几种基于过渡区提取的图象分割算法进行了较为全面的综合评价,包括分割质量评价、抗噪性评价、运算复杂度评价、目标背景面积比影响及人为参数等方面的评价,评价使用了4种典型的过渡区提取与分割算法,以及两种经典的阈值化分割算法。评价结果表明,过渡区提取与分割算法可以取得优于传统阈值化分割算法的综合性能,是一种值得研究的图象分割方法。
|
| 关 键 词： | 图象分割 分割评价 过渡区 |
| 文章编号： | 1002-8331-(2005)08-0060-04 |
Evaluation on Transition Region Extraction Based Image Segmentation |
| |
| Yan Chengxin,Sang Nong,Zhang Tianxu. Evaluation on Transition Region Extraction Based Image Segmentation[J]. Computer Engineering and Applications, 2005, 41(8): 60-63 |
| |
| Authors: | Yan Chengxin Sang Nong Zhang Tianxu |
| |
| Affiliation: | Yan Chengxin1 Sang Nong2 Zhang Tianxu21 |
| |
| Abstract: | Compositive evaluations are made on transition region extraction based image segmentation methods.Evaluating aspects include segmentation quality,ability to deal with noise,computation speed,effects of object/background ratio,controlling parameters and so on.Segmentation methods include four typical transition region extraction and segmentation approaches,and two traditional thresholding approaches.Experimental results demonstrate that transition region extraction based image segmentation methods can obtain better compositive evaluation performance than traditional thresholding method,and transition region extraction based image segmentation is an promising method to be further studied. |
| |
| Keywords: | image segmentation segmentation evaluation transition region |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
