基于多级分类的+PojSAR+图像机场跑道检测.pdf

第36卷第5期系统工科与电子技术Vol 36 No 52014年5月ystems Engineering and Electronics文章编号:1001-506X(20141)05086606网址:www.sys-ele.con基于多级分类的 POSAR图像机场跑道检测韩萍,徐建飒,赵爱军2(1.中国民航大学天津市智能信号与图像处理重点实验室,天300300东方通用航空有限责任公司,天津300300)摘要:提出一种复杂场景极化合成孔径雷达( polarimetric synthetic aperture radar, POISAR)图像机场跑道多级分类检测方法。首先利用先验信息进行q第一级分类,得到图像中各类训练样本模板;然后利用 POLSAR图像极化相干矩阵的统计特性进行第二级分类;再根据跑道的弱回波特性,利用极化总功率检测器完成第三级分类,提取图像中疑似机场跑道区域:最后根据机场跑道的尺寸及结构特征进行判别,确定机场跑道区域。采用美国UAVS.AR机載系统获取的多组实测測数据对算法进行验证?并与现有的两种方法进行比较,結果表明,本文算法能有效地检測出跑道,并保持跑道结构完整、轮廓清晰且虛警率低。关键词:极化合成孔径雷达图像;复杂场景;跑道检测;/q分类;多级分类中图分类号:TP753文献标志码:ADOI:10.3969/j.1issn.1001-506X.2014.05.09POISAR image runways detection based on multi-stage classificationHAN Ping , XU Jian-sa, ZHAO Ai-jun(1. Tianjin Key Lab for Intelligent Signal and Image Processing, Civil Awation University of ChinaTianjin 00.300, China; 2. Easiern General Awialion Corporation Limited, Tianjin 300.00. ChinaAbstract: A new effective algorithm for complex scenes of polarimetric synthetic aperture radar (POLSAR)image runway detection based on multi-stage classification is developed. Firstly, training sample templates areselected according to h/q classification with the priori of backward scattering mechanism of the terrain. Thenthe statistical characteristic of polarized coherent matrix of POLSAR images are used in the second stage classifcation. Thirdly, according to weak backscattering characteristic of the runways, the third stage classification isdone by using the span detector. Finally, the airport runway region is identified with the size and structuracharacleristics of the runways. The datasets obtained by the U. S. UAVSAR airborne systems are used to dothe experiments. Results show that comparing with the existing two methods, the new algorithm can detect therunways effectively with intact structures, clear outlines and low false alarm ratesKeywords: polarimetric synthetic aperture radar (POLSAR) image; complex scenc; runway detection; h/qclassification multi-stage classification0引言射特性,困此,对其在各个领域的应用研究也成为各国学者研究的热点机场是笨用和民用的重要设施,其自动检测被厂广泛应日前,由于技术保密的需要,国内外可阅的有关极化用于军事侦察、紧急救援、飞机檢测等众多领域。现有的关合成孔径雷达( polarimetric synthetic aperture radar.Pol于机场跑道检测的研究主要是针对红外图像、可见光遥感SAR)图像机场跑道检测的文献较少。近具代表性的图像以及合成孔径雷达( synthetic aperture radar.研究成果有文献10-11。文献10主要采用独立分量分SAR)图像-12,SAR图像因其获取不受大气、光照等条件析以及序列非线性滤波提取机场跑道,较适用于具有强烈的限制,被广泛应用于各类日标檢测领域。相比于単柲化相干斑点噪声的多柲化SAR影像处理;文献111先采用灰SAR?像,全枚化SAR像可以更好地表征口标的极化散度和柲化度两个极化特量增强图像,再利用水平集模型收稿日期:2013-06-09;修回日期:2014-01-27;网络优先出版日期:2011-03-14网络优先出版地址:hp:/www.cnki.ne/kcms/detail/1.2422.TN.20140314.1346.001.huml基金项目:国家自然科学基金重点项日(61231017);中央高校基本科研业务费安项基金(3122014C004);中国民航人学科研基金(2012KYEO3)资助课题