Robust Detection of Single and Double Persistent Scatterers in Urban Built Environments
Robust Detection of Single and Double Persistent Scatterers in Urban Built Environments
复制标题
DOI：
10.1109/tgrs.2015.2496193
复制
发表时间：
2016-04
影响因子：
8.2
通讯作者：
Peifeng Ma;Hui Lin
中科院分区：
文献类型：
--
作者：
Peifeng Ma;Hui Lin
在本文中，我们开发了一种合成孔径雷达（SAR）的层析成像为基础的持久性散射体干涉（Tomo-PSInSAR）方法来检测城市建成环境中的单，双持久性散射体（PS）。通过构造一个双层网络，我们可以联合检测单，双PS，而无需在整个区域的大气相屏（APS）的初步删除。该技术更适用于高层建筑环境（例如，香港）多云及有雨天气，在移除APS时存在很大的不确定性。在第一层网络中，我们的目标是通过构建Delaunay三角网来检测最可靠的单PS（SPS）。为了提高估计的鲁棒性，我们将联合收割机波束形成与M估计器结合起来进行弧段参数估计，并引入岭估计器进行网络调整。在第二层网络中，我们通过构建本地星星网络，使用在第一层网络中检测到的SPS作为参考点，检测剩余的SPS和所有的双PS（DPS）。为了简化DPS的检测，我们采用局部最大值比率方法来提取重叠的DPS。最后，利用56幅香港TerraSAR-X影像对Tomo-PSInSAR方法进行了验证。
In this paper, we develop a synthetic aperture radar (SAR) tomography-based persistent scatterer interferometry (Tomo-PSInSAR) method to detect single and double persistent scatterers (PSs) in urban built environments. By constructing a two-tier network, we can jointly detect single and double PSs with no need for preliminary removal of the atmospheric phase screen (APS) in the whole area. This technique is more applicable in high-rise built environments (e.g., Hong Kong) with cloudy and rainy weather, where there is much uncertainty when removing the APS. In the first-tier network, we aim to detect the most reliable single PSs (SPSs) by constructing a Delaunay triangulation network. To improve the robustness of estimation, we combine beamforming with an M-estimator for parameter estimation at the arcs and introduce a ridge estimator for network adjustment. In the second-tier network, we detect the remaining SPSs and all of the double PSs (DPSs) by constructing local star networks that use the SPSs detected in the first-tier network as reference points. To simplify the detection of DPSs, we employ a local maximum ratio method for extracting overlaid DPSs. Finally, 56 Hong Kong TerraSAR-X images are used to validate the Tomo-PSInSAR method.