获取震后建筑物震害信息有利于开展人员救援和灾后重建工作。由于高分辨率合成孔径雷达(SAR)数据少有震前数据存档,利用震后单时相高分辨率SAR数据评估建筑物震害成为研究热点,但利用高分辨率SAR数据对单体建筑物的研究却很少。以三景...获取震后建筑物震害信息有利于开展人员救援和灾后重建工作。由于高分辨率合成孔径雷达(SAR)数据少有震前数据存档,利用震后单时相高分辨率SAR数据评估建筑物震害成为研究热点,但利用高分辨率SAR数据对单体建筑物的研究却很少。以三景北川老县城震后0.24 m TerraSAR-X聚束模式(ST)数据为数据源,经多视处理后提取建筑物纹理特征,对比分析不同视数大小和纹理计算窗口大小对建筑物震害识别影响,确定最佳纹理计算窗口大小和视数大小。结合震前光学数据,获得SAR单体建筑物轮廓图,随机选取建筑物轮廓样本作为训练样本,引入支持向量机(SVM)和随机森林(RF)分类器识别建筑物震害信息。结果表明,基于纹理特征的SVM、RF方法能有效地识别高分辨SAR影像单体建筑物震害信息,SVM识别精度均在80%以上,RF识别精度均在83%以上。可见基于高分辨率SAR数据的纹理特征识别建筑物震害方法稳定有效,可为灾后应急、灾害评估和灾后重建工作提供可靠信息支撑。展开更多
D-InSAR is currently one of the most popular research tools in the field of Microwave Remote Sensing. It is unrivaled in its aspect of measuring ground deformation due to its advantages such as high resolution,continu...D-InSAR is currently one of the most popular research tools in the field of Microwave Remote Sensing. It is unrivaled in its aspect of measuring ground deformation due to its advantages such as high resolution,continuous spa-tial-coverage and dynamics. However,there are still a few major problems to be solved urgently as a result of the intrin-sic complexity of this technique. One of the problems deals with improving the accuracy of measured ground deforma-tion. In this paper,various factors affecting the accuracy of ground deformation measured by D-InSAR are systemati-cally analyzed and investigated by means of the law of measurement error propagation. At the same time,we prove that the ground deformation error not only depends on the errors of perpendicular baselines as well as the errors of the inter-ferometric phase for topographic pair and differential pair,but also on the combination of the relationship of perpen-dicular baselines for topographic pairs and differential pairs. Furthermore,a feasible approach for improving the accu-racy of measured ground deformation is proposed,which is of positive significance in the practical application of D-InSAR.展开更多
文摘获取震后建筑物震害信息有利于开展人员救援和灾后重建工作。由于高分辨率合成孔径雷达(SAR)数据少有震前数据存档,利用震后单时相高分辨率SAR数据评估建筑物震害成为研究热点,但利用高分辨率SAR数据对单体建筑物的研究却很少。以三景北川老县城震后0.24 m TerraSAR-X聚束模式(ST)数据为数据源,经多视处理后提取建筑物纹理特征,对比分析不同视数大小和纹理计算窗口大小对建筑物震害识别影响,确定最佳纹理计算窗口大小和视数大小。结合震前光学数据,获得SAR单体建筑物轮廓图,随机选取建筑物轮廓样本作为训练样本,引入支持向量机(SVM)和随机森林(RF)分类器识别建筑物震害信息。结果表明,基于纹理特征的SVM、RF方法能有效地识别高分辨SAR影像单体建筑物震害信息,SVM识别精度均在80%以上,RF识别精度均在83%以上。可见基于高分辨率SAR数据的纹理特征识别建筑物震害方法稳定有效,可为灾后应急、灾害评估和灾后重建工作提供可靠信息支撑。
基金Projects 400471090 supported by the National Natural Science Foundation of China and 1421 by the European Space Agency
文摘D-InSAR is currently one of the most popular research tools in the field of Microwave Remote Sensing. It is unrivaled in its aspect of measuring ground deformation due to its advantages such as high resolution,continuous spa-tial-coverage and dynamics. However,there are still a few major problems to be solved urgently as a result of the intrin-sic complexity of this technique. One of the problems deals with improving the accuracy of measured ground deforma-tion. In this paper,various factors affecting the accuracy of ground deformation measured by D-InSAR are systemati-cally analyzed and investigated by means of the law of measurement error propagation. At the same time,we prove that the ground deformation error not only depends on the errors of perpendicular baselines as well as the errors of the inter-ferometric phase for topographic pair and differential pair,but also on the combination of the relationship of perpen-dicular baselines for topographic pairs and differential pairs. Furthermore,a feasible approach for improving the accu-racy of measured ground deformation is proposed,which is of positive significance in the practical application of D-InSAR.
