高分辨SAR相干系数图像的开采沉陷区提取

Extraction of mining subsidence area based on coherence image of high resolution SAR

针对煤矿区开采沉陷下沉迅速、地表变形大、容易形成低相干区的特点,该文采用神东矿区2012年1月至2012年12月RADARSAT-2 MF5多视精细波束模式图像,经相干性计算、对低相干区进行数字图像处理、人机交互解译等过程获取研究区内沉陷区的数量和规模。将基于InSAR相干系数图像提取的沉陷区与DInSAR结果进行叠加对比,空间位置能够很好地吻合,由此获得连续变化的开采沉陷区39处。实验表明:不需要复杂的干涉测量,仅使用相干系数图像也可以提取开采沉陷区,可以实现大范围、高强度开采活动造成的地表沉陷灾害的快速评估。

In view of some characteristics of Shendong coal mining subsidence,such as rapid subsidence,large-scale deformation,and easily lead to low coherence or incoherence,in this paper,agroup of multi-fine beam mode images from RADARSAT-2 are used during Jan.2012 to Dec.2012.By means of the coherence calculation,digital image processing,human-computer interactive interpretation,a certain size and number of subsidence areas are obtained.Comparison the subsidence area extracted by the InSAR coherence image with the result by DInSAR,the space positions of them can be well matched,so that 39 continuous changes mining subsidence area were obtained.Experiments show that the coherent coefficient images can also be used to extract mining subsidence areas without the need for complex interferometric measurements,which will enable rapid assessment of surface subsidence disasters caused by large-scale,high-intensity mining activities.