| 一种快速模糊聚类分割算法 |
| |
| 引用本文： | 李爱生,黄铁侠,柳健. 一种快速模糊聚类分割算法[J]. 华中科技大学学报(自然科学版), 1992, 0(4) |
| |
| 作者姓名： | 李爱生 黄铁侠 柳健 |
| |
| 作者单位： | 武汉通信指挥学院(李爱生)，华中理工大学电子与信息工程系(黄铁侠)，华中理工大学电子与信息工程系(柳健) |
| |
| 摘 要： | 本文把区域生长技术与FCM聚类方法结合起来,提出了一种快速FCM聚类分割算法.由于大大减少了参与聚类的样本数目,有效地提高了FCM聚类分割的速度.通过对遥感TM图像的分割实验,本算法比经典FCM聚类算法速度提高三倍以上.
|
| 关 键 词： | 遥感图象分割 FCM(模糊C均值)聚类 区域生长 |
A Fast Fuzzy Clustering Segmentation Algorithm |
| |
| Li Aisheng Huang Tiexia Liu Jian. A Fast Fuzzy Clustering Segmentation Algorithm[J]. JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE, 1992, 0(4) |
| |
| Authors: | Li Aisheng Huang Tiexia Liu Jian |
| |
| Affiliation: | Li Aisheng Huang Tiexia Liu Jian |
| |
| Abstract: | With region-growing techniques of the new algorithm and the FCM (Fuzzy C-Means) clustering method, a fast FCM clustering segmentation algorithm is developed. It has two steps, initial segmentation of images with the simple region-growing algorithm and clustering segmentation of images with FCM clustering algorithm after initial segmentation. The initial segmentation is aimed at reducing the sample number of FCM clustering so as to effectively increase the convergence rate of FCM clustering segmentation. The algorithm presented asks for less memory capacity and is at least three times faster in convergence rate than that of the classical FCM clustering algorithm. This algorithm has been used to segment TM (Thematic Mapper) images of the city proper of Wuhan and its feasibility and reliability are proved. |
| |
| Keywords: | segmentation of remotely sensed images FCM clustering region- growing |
| 本文献已被 CNKI 等数据库收录！ |
