结合凝聚层次聚类的极化SAR海冰分割

于波; 孟俊敏; 张晰; 纪永刚

doi:10.11834/jrs.20132091

遥感应用 | 浏览量 : 0 下载量: 316 CSCD: 4 更多指标

PDF
导出
分享
收藏
专辑

结合凝聚层次聚类的极化SAR海冰分割
Segmentation method for agglomerative hierarchical-based sea ice types using polarimetric SAR data
2013年17卷第4期页码：887-904
纸质出版日期： 2013 ，
DOI： 10.11834/jrs.20132091

扫描看全文

[1]于波,孟俊敏,张晰,纪永刚.结合凝聚层次聚类的极化SAR海冰分割[J].遥感学报,2013,17(04):887-904.

Segmentation method for agglomerative hierarchical-based sea ice types using polarimetric SAR data[J]. Journal of Remote Sensing, 2013,17(4):887-904.
[1]于波,孟俊敏,张晰,纪永刚.结合凝聚层次聚类的极化SAR海冰分割[J].遥感学报,2013,17(04):887-904. DOI： 10.11834/jrs.20132091.

Segmentation method for agglomerative hierarchical-based sea ice types using polarimetric SAR data[J]. Journal of Remote Sensing, 2013,17(4):887-904. DOI： 10.11834/jrs.20132091.

摘要

针对全极化高分辨率合成孔径雷达（SAR）多噪声的特点

选择合适的海冰SAR影像纹理特征量

并基于凝聚层次聚类的思想

提出一种全极化SAR海冰分割方法。将该方法的结果与K-Means、迭代自组织聚类（ISODATA）和模糊C均值算法等经典分割方法相比较

发现该方法的碎斑明显减少

且分割结果较为准确

证明了该方法的有效性。

Abstract

This paper investigates the characteristics of high resolution full-polarization Synthetic Aperture Rader （ SAR） data combined with the selected texture characteristic

and uses the theory of the agglomerative hierarchical clustering algorithm to present a sea ice segmentation method using full-polarimetric SAR data. By comparing the result of this method with the results of the K-Means

ISODATA

and fuzzy c-means methods

we verified that our method can reduce the number of small patches and realize good segmentation accuracy. The results verify the validity of this method.