先验信息驱动的DSInSAR自适应相位优化算法  

The adaptive phase optimization algorithm for DSInSAR driven by priori information

作　　者：李世金 卞正富[1,2] 高延东 张书毕[1,2] 郑南山[1,2] 张秋昭[1,2] 张艳锁 田雨 LI Shijin;BIAN Zhengfu;GAO Yandong;ZHANG Shubi;ZHENG Nanshan;ZHANG Qiuzhao;ZHANG Yansuo;TIAN Yu(School of Environment and Spatial Informatics,China University of Mining and Technology,Xuzhou,Jiangsu 221l16,China;Key Laboratory of Land Environment and Disaster Monitoring of Natural Resources,China University of Mining and Technology,Xuzhou,Jiangsu 221l16,China)

机构地区：[1]中国矿业大学环境与测绘学院,江苏徐州221116 [2]中国矿业大学自然资源部国土环境与灾害监测重点实验室,江苏徐州221116

出　　处：《中国矿业大学学报》2024年第2期409-420,共12页Journal of China University of Mining & Technology

基　　金：中央高校基本科研业务费专项资金项目(2023QN1043)。

摘　　要：相位优化是分布式散射体干涉合成孔径雷达(Distributed Scatterers Interferometric Synthetic Aperture Radar,DSInSAR)技术中提升相位信噪比的关键步骤.为了改善现有相位优化算法在矿区大梯度形变对应密集条纹处的相位信息严重损失问题,提出一种基于先验信息驱动的DSInSAR自适应相位优化算法.算法首先采用常规小基线集技术反演初始时序形变相位,通过时序相位预处理获取先验形变相位信息;然后将其从原始单视复数(Single Look Complex,SLC)相位中去除,获取SLC残余相位,结合相干性次幂加权策略构建残余相位优化模型,通过先验形变相位补偿,估计最终优化相位.试验结果表明:提出算法可有效兼顾大梯度形变场对应条纹密集区域的相位信息保护及噪声抑制,且有效提升了大梯度形变区域的监测点密度及监测精度,较常规优化算法具有更好的自适应效果及优化性能.Phase optimization is a key step in distributed scatterers interferometric synthetic aperture radar(DSInSAR)technique to improve the phase signal-to-noise ratio.In order to improve the severe loss of phase information at dense fringes corresponding to large gradient deformation in mining areas using current phase optimization algorithms,an adaptive phase optimization algorithm for DSInSAR driven by a priori information is proposed.The proposed algorithm first uses the conventional small baseline subset InSAR technique to invert the initial time series deformation phase,and obtains a priori deformation phase information through the preprocessing for the time series deformation phase.Then,the priori deformation phase is removed from the original single look complex(SLC)phase to obtain the SLC residual phase,and the residual phase optimization model is constructed by combining the coherence-power-weighting strategy.Furthermore,the final optimized phase is estimated by compensating for a priori deformation phase.The experimental results show that the proposed algorithm can effectively take into account the phase information protection and noise suppression in the dense fringe region corresponding to the large gradient deformation field,and effectively improve the measurement point density and monitoring accuracy in the large gradient deformation region.In summary,the proposed algorithm has better adaptive effect and optimization performance than the conventional optimization algorithm.

关 键 词：DSInSAR 相位优化 先验信息 矿区形变监测 

分 类 号：P237[天文地球—摄影测量与遥感]