论文部分内容阅读
针对震前震后合成孔径雷达(SAR)图像中发生复杂形变的目标,提出了基于稳健的加权核主成分分析(KPCA)的配准方法。首先,提出一种稳健的加权KPCA(RWKPCA)方法,不仅能获得震前震后形变目标的共同稳健核主成分(RKPCs),还可以作为异常值判别准则;其次,利用在共同RKPCs上的投影定义震前震后形变目标特征的相似性度量;最后,利用特征的相似性度量精确配准形变目标。对2008年5月12日汶川地震前后的SAR图像进行配准并与现有方法进行比较,结果表明,本文方法能够有效的得到形变目标的共同RKPCs,并得到很好的配准结果。
Aiming at the target of complex deformation in the synthetic aperture radar (SAR) image after the pre-earthquake earthquakes, a registration method based on robust weighted principal component analysis (KPCA) is proposed. First of all, a robust weighted KPCA (RWKPCA) method is proposed, which not only can obtain common robust kernel principal components (RKPCs) of pre-earthquake post-earthquake deformation targets, but also can be used as criterion of outlier detection. Secondly, using the projection definition on common RKPCs Pre-earthquake earthquake target feature similarity measure; Finally, the use of feature similarity measure accurate registration of deformation targets. The registration of SAR images before and after Wenchuan earthquake on May 12, 2008 is compared with the existing methods. The results show that the proposed method can effectively obtain the common RKPCs of deformation targets and obtain good registration results.