改进的遗传算法用于极化合成孔径雷达影像非监督分类

doi:10.11873/j.issn.1004-0323.2013.1.65

遥感技术与应用

2013, Vol. 28

Issue (1): 65-71 DOI: 10.11873/j.issn.1004-0323.2013.1.65

图像与数据处理

改进的遗传算法用于极化合成孔径雷达影像非监督分类

王刚^2,4,余洁^1,2,朱腾²,张中山^2,3,赵争⁵

(1.首都师范大学资源环境与地理信息系统北京市重点实验室,北京 100048;2.武汉大学遥感信息工程学院,湖北武汉 430079;3.中国电子科技集团第38研究所,安徽合肥 230031;4.青岛市勘察测绘研究院,山东青岛 266031;5.中国测绘科学研究院,北京 100039)

Researsh on Polarimetric SAR Image Unsupervised Classification based on Improved Genetic Algorithm

Wang Gang^2,4,Yu Jie^1,2,Zhu Teng²,Zhang Zhongshan^2,3,Zhao Zheng⁵

(1.Beijing Key Lab of Resources Environment and Geographic Information System,
Capital Normal University,Beijing 100048,China;
2.School of Remote Sensing and Information Engineering,Wuhan University,Wuhan 430079,China;
3.The 38th Research Institute,China Electronic Technology Group,Hefei 230031,China;
4.Institute of Surveying and Mapping,Qingdao 266031,China;
5.Chinese Academy of Surveying and Mapping,Beijing 100039,China)

全文: PDF(9233 KB)

摘要：

H/α-Wishart分类方法是目前常用且较为有效的极化SAR影像分类方法,但其分类精度还有待改善。研究一种基于遗传算法的极化SAR影像的分类方法,该方法根据极化SAR影像Cloude特征分解的特征值,采用H/α平面进行初分类,然后采用遗传算法迭代进行再次分类。针对遗传算法“早熟”和收敛速度慢的问题,结合H/α平面图对遗传算法的变异算子进行了改进,以利用极化散射机理缩小变异范围,改善算法收敛速度。采用NASA-JPL实验室的极化SAR数据以及中国电子科技集团38研究X波段原型样机的高分辨率极化SAR数据进行实验,结果表明:该方法极化SAR影像分类精度优于H/α-Wishart分类方法。

关键词： 遗传算法; 变异算子; 极化合成孔径雷达; 非监督分类

Abstract:

The H/a-Wishart classifier is a kind of common and reliable classification methods for the polarimetric SAR(PolSAR) image,of which the classification precision still needs to be improved.A new classification algorithm for the PolSAR image is introduced in this paper,which is based on the genetic algorithm.This method decomposes polarimetric SAR data to extract features,then initials classification by the H/α plane and finally does reclassification by iterating the genetic algorithm.The disadvantages of genetic algorithm are‘premature’ and slow convergence.This paper improves the mutation operator of genetic algorithm by combining the H/α plane so that the algorithm can make use of the polarimetric scattering mechanism to decrease the searching field and improve the efficiency.Adopting the data of the NASA-JPL Laboratory and high resolution PolSAR data of The 38th Research Institute of China Electronic Technology Group to research,the results show that the refined genetic algorithm can improve the accuracy of classification and get better results than both the H/a-Wishart classifier and standard genetic algorithm.

Key words: Genetic algorithm Mutation operator Polarimetric SAR Unsupervised classification

收稿日期: 2011-11-14 出版日期: 2013-06-21

TP 79

基金资助:

国家863计划项目(2011AA120404)。

通讯作者: 余洁(1964-),女,湖南长沙人,教授,主要从事遥感与GIS科研和教学工作。Email：yuj2011@whu.edu.cn。

作者简介: 王刚(1986-),男,山东青岛人,硕士研究生,主要从事极化SAR影像处理研究。Email:327188204@qq.com。

	服务
	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	王刚
	余洁
	朱腾
	张中山
	赵争

引用本文:

王刚,余洁,朱腾,张中山,赵争. 改进的遗传算法用于极化合成孔径雷达影像非监督分类[J]. 遥感技术与应用, 2013, 28(1): 65-71.

Wang Gang,Yu Jie,Zhu Teng,Zhang Zhongshan,Zhao Zheng. Researsh on Polarimetric SAR Image Unsupervised Classification based on Improved Genetic Algorithm. Remote Sensing Technology and Application, 2013, 28(1): 65-71.

链接本文:

http://www.rsta.ac.cn/CN/10.11873/j.issn.1004-0323.2013.1.65 或 http://www.rsta.ac.cn/CN/Y2013/V28/I1/65

［1］Cloude S R,Pottier E.An Entropy based Classification Sch-eme for Land Applications of Polarimetric SAR［J］.IEEE Transactions on Geoscience and Remote Sensing,1997,35(1):68-78.

［2］Lee J S,Grunes M R,Ainsworth T L,et al.Unsupervised Cl-assification Using Polarimetric Decomposition and the Complex Wishart Classifier［J］.IEEE Transactions on Geoscience and Remote Sensing,1999,37(5):2244-2266.

［3］Yang Shuying.Image Pattern Recognition-VC++ Technology to Achieve［M］.Beijing:Tsinghua University Press,2005.［杨淑莹.图像模式识别-VC++技术实现［M］.北京:清华大学出版社,2005.］

［4］Ferro-famil L,Pottier E,Lee J S.Unsupervised Classfication of Multifrequency and Fully Polarimetric SAR Images based on the H/A/Alpha-Wishart Classifier ［J］.IEEE Transactions on Geoscience and Remote Sensing,2001,39(11):2300-2355.

［5］Ge Peiming.Improved Genetic Algorithms and Applications in Engineering Optimization［D］.Chengdu:Southwest Jiaotong University,2006.［葛培明.改进的遗传算法及其在工程优化中的应用［D］.成都:西南交通大学,2006.］

［6］Ferro-famil L,Pottier E,Lee J S.Unsupervised Classfication 〖JP2〗of Multifrequency and Fully Polarimetric SAR Images based on the H/A/Alpha-Wishart Classifier［C］//IEEE Transactions on Geoscience and Remote Sensing,2001,39(11):2300-2355.

［7］Lin W,Tian Z,Wen X B,et al.Unsupervised Segmentation of the Pol SAR Image Using Similarity Parameters in Sequential Projection Pursuit Model［C］//International Conference on Signal Processing Proceedings,Beijing,China,2004:772-774.

［8］Chen Zhong,Liu Jianguo,Wang Guoyou.Research of Unsupervised Classification of Remote Sensing Image based on Modified Genetic Algorithm ［J］.Computer Engineering and Applications,2008,44(19):78-79.［陈忠,刘建国,汪国有.基于改进遗传算法遥感图像非监督分类研究［J］.计算机工程与应用,2008,44(19):78-79.］

［9］Lee K Y,Liew S C,Kwoh L K.Land Cover Classification of Polarimetric Synthetic Aperture Radar (POLSAR) Data based on Scattering Mechanisms and Complex WishartDistribution［C］//International Geoscience and Remote Sensing Symposium,Toronto,Canada,2002,6(5):2602-2604.

［10］Ainsworth T L,Lee J S,Chang L W.Classification Comparisons Between Dual-Pol and Quad-Pol SAR Imagery［C］//ISPRS Journal of Photogrammetry and Remote Sensing,2007,64(5):464-471.

［11］Zhou X G,Zhao L W,Kuang G Y,et al.Unsupervised Classification of PolSAR Images Using Eigenvector Analysis,Krogager Decomposition and the Wishart Classifier［C］//Proceedings of the 2008 IEEE International Conference on Information and Automation,Hunan,China,2008,6:161-164.

［12］Tan L L,Yang R L.Unsupervised Classification of PolSAR Data Using Freeman Decomposition and Fuzzy Clustering［C］//Asia-Pacific Conference on Synthetic Aperture Radar,Xian,China,2009,10:489-493.

[1]	姜萍，刘修国，陈启浩，邵芳芳. 利用多尺度SVM-CRF模型的极化SAR图像建筑物提取[J]. 遥感技术与应用, 2017, 32(3): 475-482.
[2]	史飞飞,高小红，杨灵玉，何林华，贾伟. 基于HJ-1A高光谱遥感数据的湟水流域典型农作物分类研究[J]. 遥感技术与应用, 2017, 32(2): 206-217.
[3]	袁春琦,徐佳,程圆娥,陈媛媛,许康. 基于协同训练与集成学习的极化SAR图像半监督分类[J]. 遥感技术与应用, 2017, 32(2): 380-385.
[4]	徐旭,张风丽,王国军,符喜优,沙敏敏,李志坤,邵芸,陈龙永,梁兴东. 基于高亮特征匹配的双视向SAR图像建筑物高度提取[J]. 遥感技术与应用, 2016, 31(1): 149-156.
[5]	沈掌泉,虞舟鲁. 基于单亲遗传算法和子像元/像元空间吸引模型的亚像元制图研究[J]. 遥感技术与应用, 2015, 30(1): 129-134.
[6]	梁军民,李立钢. 一种应急救灾飞艇部署与规划方法研究[J]. 遥感技术与应用, 2014, 29(4): 639-645.
[7]	吴健生,林倩,李卫锋,刘建政,彭建. 基于特征选择的土地整理地物识别规则的动态提取[J]. 遥感技术与应用, 2013, 28(5): 799-806.
[8]	魏铼,胡卓玮. SAR影像海洋表面溢油检测方法与实现[J]. 遥感技术与应用, 2013, 28(1): 34-43.
[9]	王庆,曾琪明,焦健. 全极化SAR极化特征谱应用研究[J]. 遥感技术与应用, 2012, 27(6): 896-903.
[10]	薛星宇,刘红玉. 基于ALOS影像的盐城海滨湿地遥感信息分类方法研究[J]. 遥感技术与应用, 2012, 27(2): 248-255.
[11]	张露, 郭华东, 李新武. 利用POLSAR数据探索极化相关系数在居民地提取中的作用[J]. 遥感技术与应用, 2010, 25(4): 474-479.
[12]	王翔宇, 汪西莉. 结合灰色扩充的GA—BP神经网络模型在渭河水质遥感反演中的应用[J]. 遥感技术与应用, 2010, 25(2): 251-256.
[13]	谢屹鹏, 汪西莉. 基于GA-LSSVR的渭河水质参数遥感反演研究[J]. 遥感技术与应用, 2010, 25(2): 257-262.
[14]	邹斌,张腊梅,孙德明,张晔. PolSAR图像信息提取技术及应用的发展[J]. 遥感技术与应用, 2009, 24(3): 263-273.
[15]	张露,郭华东,廖静娟,韩春明. 包敦其河冲积扇全极化SAR数据同极化相关特征分析[J]. 遥感技术与应用, 2008, 23(4): 385-388.

Viewed

Full text

Abstract

Cited

Shared

Discussed