基于PWF的极化SAR图像匹配方法  

Image Matching Method for Polarimetric SAR Based on PWF

作　　者：马文婷[1] 曲迪[2] 吴蔚[1] 熊朝华[1] 秦洪[1] 

机构地区：[1]信息系统工程重点实验室,南京210007 [2]中国电子科技集团公司第二十八研究所,南京210007

出　　处：《指挥信息系统与技术》2016年第2期63-67,共5页Command Information System and Technology

摘　　要：极化合成孔径雷达(SAR)图像匹配是修正惯性导航累积误差和提高导航精度的重要手段。由于极化SAR图像的斑点噪声降低了目前多数匹配方法性能,故提出了一种基于极化白化滤波器(PWF)的极化SAR图像匹配方法,利用极化信息抑制斑点噪声从而改善了匹配效果。运用该方法和经典的极化总功率(SPAN)+尺度不变特征变换(SIFT)方法,对多组不同波段的SAR图像进行匹配试验,结果表明该方法在匹配点数、精度和成功率上均优于SPAN+SIFT匹配方法。The image matching for polarimetric synthetic aperture radar（SAR）plays an important role in correcting the accumulative errors of inertial navigation system（INS）,thus enhancing the navigation precision.Because the performance of most matching methods is degraded by the speckle noises in polarimetric SAR images,the image matching method for polarimetric SAR based on the polarimetric whiten filter（PWF）is proposed.The method can suppress the speckle noises with the polarimetric information,thus the matching effect is improved.The matching experiments are performed among groups of SAR images in different bands with the method and the classical method integrated the methods of the total polarimetric power parameter SPAN ＋scale invariant feature transform（SIFT）.Results show that the method can obtain better performance on the count of matching dots,the matching precision and the matching probability than the classical SPAN ＋ SIFT method.

关 键 词：极化合成孔径雷达图像 图像匹配 斑点噪声 极化白化滤波器 尺度不变特征变换 

分 类 号：TN959.21[电子电信—信号与信息处理]