基于样本正交子空间的SAR目标识别方法

韩征; 苏志刚; 韩萍; 吴仁彪

doi:10.3724/SP.J.1146.2008.01510

基于样本正交子空间的SAR目标识别方法

doi: 10.3724/SP.J.1146.2008.01510

基金项目:

国家自然科学基金(60736009, 60872110)和国家863计划项目(2006AA12Z321)资助课题

计量
- 文章访问数: 3319
- HTML全文浏览量: 87
- PDF下载量: 863
- 被引次数: 0
出版历程
- 收稿日期: 2008-11-14
- 修回日期: 2009-04-06
- 刊出日期: 2009-11-19

SAR Target Recognition Method Based on Orthogonal Subspace of Samples

摘要

摘要: 利用合成孔径雷达(Synthetic Aperture Radar, SAR)获取的目标像进行识别时，基于子空间的自动目标识别(Automatic Target Recognition, ATR)方法通常是对样本数据的值空间进行操作。当识别相似目标时，由于彼此的值空间存在较大的交集，生成的识别模板的可分性较差。该文提出一种SAR目标ATR方法，该方法将SAR样本数据的正交子空间作为投影空间。因此，不同类目标样本在投影空间的差异性加大，能够明显提高识别效果。实验结果表明，本文方法的识别性能优于其它同类方法。
- 自动目标识别; 合成孔径雷达; 特征模板; 正交子空间
Abstract: To identify with the target images obtained by Synthetic Aperture Radar (SAR), those subspace-based methods for Automatic Target Recognition (ATR) are usually based on the range subspace of the samples. When some similar targets need to be distinguished, the corresponding templates are poorly separable because the range subspaces of each other have a big intersection. A method for SAR target ATR is proposed in this paper, which considers the orthogonal complement subspace of the samples as the projection subspaces. Consequently, the difference, between the projections of the different type targets on the projection subspace, is enlarged, which significantly improves the identification performance. Experimental results show that the proposed method is superior to the other similar methods.

HTML全文

参考文献(1)

Sun Yi-jun, Liu Zhi-peng, and Todorovic Sinisa, et al..Synthetic aperture radar automatic target recognition usingadaptive boosting. SPIE The International Society forOptical Engineering, San Diego, 2005, 5: 282-293.[2]Yang Yi-nan, Qiu Yu-xia, and Lu Chao. Automatic targetclassification experiments on the MSTAR SAR images.Proceedings of Sixth International Conference onSNPD/SAWN, Towson, 2005: 2-7.[3]Duan Rong, Chen Ling, and Man Hong . Adaptive SAR ATRin likelihood space. Algorithms for Synthetic Aperture RadarImagery XII, Proceedings of SPIE 2005, 5808: 294-305.[4]Wissinger J, Washburn R, and Morgon D, et al.. Searchalgorithms for model-based SAR ATR. Algorithms forSynthetic Aperture Radar Imagery III, Proceedings of SPIE1996, Orlando, FL, 1996, 2757: 279-293.[5]Novak L M, Owirka G J, and Brower W S, et al.. Theautomatic target recognition system in SAIP. The LincolnLaboratory Journal, 1997, 10(2): 185-202.[6]Potter L C and Moses R L. Attributed scattering centers forSAR ATR[J].IEEE Transactions on Image Processing.1997,6(1):79-91[7]高贵, 计科峰, 匡纲要, 等. 高分辨率SAR 图像目标峰值提取及峰值稳定性分析. 宇航学报, 2004, 25(1): 102-108.Gao Gui, Ji Ke-feng, and Kuang Gang-yao, et al.. Targetspeak extraction from high-resolution SAR image and peaksstability analysis. Journal of Electronics InformationTechnology, 2004, 25(1): 102-108.[8]胡利平, 刘宏伟, 吴顺君. 一种新的SAR 图像目标识别预处理方法. 西安电子科技大学学报(自然科学版), 2007, 5(34):733-737.Hu Li-ping, Liu Hong-wei, and Wu Shun-jun. Novelpre-processing method for SAR image based automatic targetrecognition. Journal of Xidian University, 2007, 5(34):733-737.[9]何壸, 刘宏伟, 吴顺君. 基于MSTAR 数据的SAR 图像自动目标识别. 2005 年中国合成孔径雷达会议, 南京, 2005:278-282.He Kun, Liu Hong-wei, and Wu Shun-jun. Automatic targetrecognition based on MSTAR SAR image database.Proceedings of CSAR-2005, Nanjing, 2005: 412-417.[10]韩萍, 吴仁彪, 王兆华, 等. 基于KPCA 准则的SAR 目标特征提取与识别[J].电子与信息学报.2003, 25(10):1297-1301浏览[11]Han Ping, Wu Ren-biao, and Wang Yun-hong, et al.. Anefficient SAR ATR approach. IEEE International Conferenceon Acoustics, Speech, and Signal Processing, Hong Kong,China, 2003, 4: 429-432.[12]王世晞, 贺志国. 基于PCA 特征的快速SAR 图像目标识别方法. 国防科技大学学报, 2008, 30(3): 137-141.Wang Shi-xi and He Zhi-guo. The fast target recognitionapproach based on PCA features for SAR images. Journal ofNational University of Defense Technology, 2008, 30(3):137-141.[13]宦若虹, 杨汝良. 基于KFD+ICA 特征提取的SAR 图像目标识别. 系统工程与电子技术, 2008, 30(7): 1237-1240.Huan Ruo-hong and Yang Ru-liang. SAR image targetrecognition based on KFD + ICA feature extraction. SystemsEngineering and Electronics, 2008, 30(7): 1237-1240.[14]Hu Li-ping, Liu Jin, and Liu Hong-wei, et al.. Automatictarget recognition based on SAR images and two-stage2DPCA features. The Asia-Pacific Conference on SyntheticAperture Radar, Huangshan, 2007, 11: 801-805.[15]Zhao Qun and Principe J C. Support vector machine for SARautomatic target recognition[J].IEEE Transactions onAerospace and Electronic Systems.2001, 37(2):643-654[16]王颖. 基于Boosting 的SAR 自动目标识别技术研究. [硕士论文], 中国民航大学, 2007.Wang Ying. Research on SAR ATR based on boosting.[Master dissertation], Civil Aviation University of China,2007.

施引文献

资源附件(0)

访问统计