【摘要】 针对目前地震灾害支援工作中,利用雷达遥感技术对灾区建筑物倒塌情况进行大范围评估时一般难以获取到震前数据的问题,提出了一种利用高分辨率干涉合成孔径雷达技术的倒塌建筑物快速评估方法。此方法采用"相干系数估计-局部倒塌识别-街区倒塌判别"3个步骤实现倒塌解译:通过采用震后干涉对作为数据源避免了缺少震前数据的问题,通过对局部影像块而非直接对像素进行处理的方式加快了倒塌评估的效率,之后使用最大投票法获取街区级倒塌程度评估结果,提高了评估精度和稳定性。以2010年玉树地震为案例,定量分析与实验结果表明该文方法在不依赖震前数据的前提下,仍可在街区级倒塌情况评估中达到较高的解译精度。更多还原

【Abstract】 Currently in the earthquake emergency rescues,the wide-range building collapse assessment using radar remote sensing technology encounters high difficulty for obtaining appropriate pre-event data.Aiming at this dilemma,a novel collapse assessment method using high-resolution interferometric synthetic aperture radar data is proposed in this paper.The method mainly implements three processing steps to achieve collapse assessment,including coherent coefficient estimation,local patch classification,and block collapse identification.Generally,the use of post-event interferometric pair avoids the dependence of preevent data,and in order to speed up the processing,the subsequent classification is based on local image patches instead of pixels.Finally,the block-level collapse assessment results can be acquired by the use of voting.In the case of Yushu earthquake in 2010,the quantitative analysis and experimental results show that,without the need of pre-event data,the proposed method can still perform well,which demonstrates its effectiveness.更多还原