一种优化的公共主影像选取方法  

An optimized method for select common master acquisition in PS-DINSAR

作　　者：刘向铜[1,2,3] 曹秋香[2] 熊助国[2] 王胜平[2] 

机构地区：[1]中南大学测绘与国土信息工程系,长沙410083 [2]江西省数字国土重点实验室,南昌330013 [3]东华理工大学,南昌330013

出　　处：《测绘科学》2016年第11期114-117,188,共5页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41204003;41304020);江西省数字国土重点实验室开放研究基金资助项目(DLLJ201415);东华理工大学省部共建核资源与环境国家重点实验室培育基地项目(NRE1508)

摘　　要：针对PS-DINSAR技术中公共主影像的选取问题,提出了基于测量平差理论的归化权值和的方法。该方法在综合相关系数方法的基础上,综合考虑了时间基线、有效空间基线和多普勒质心频率差的影响,利用平差的思想进行定权,来选择公共主影像。归化权值和的方法主影像的选择更加合理、可靠。基于ERS-1/2的影像数据进行实验的实验说明:该算法在主影像的选取上更加稳定,数据的统计特性也更优化,对PS公共主影像的优化选择有重要参考价值。The article aiming at the problem of selecting common master image in PS-DINSAR （Per- manent Scatter Synthetic Aperture Radar）, proposed the sum of normalized weights model based on the theory of Gross Error Elimination of Surveying. Based on joint correlation coefficient algorithm, consider- ing the effects of temporal baselines, spatial baselines and Doppler centroid frequency difference of image pairs, the sum of normalized weights model is proposed to select common master image using the ideas of Surveying to get out the effecting weights. The sum of normalized weights model is more reasonable and reliable. Experiment is carried out with 19 ERS- 1/2 SLC SAR images to shows that. compared with joint correlation coefficient algorithm, the mothed is more reliable and more stable in selecting common master image with more optimized statistical properties. So, the model has important reference value in selecting common master image in PS-DINSAR.

关 键 词：PS-DINSAR 公共主影像 优化选取 

分 类 号：P237[天文地球—摄影测量与遥感]