一种顾及相邻影像偏移量的基线数据链式配准方法

A Baseline Data Chain Registration Method Considering the Offset of Adjacent Images

在铁路勘察设计选线阶段,InSAR是采空区、滑坡等地质灾害探测的重要技术手段,为了解决TOPS模式下长时间基线数据容易产生相位跳变,配准精度较低等问题,提出一种顾及相邻影像偏移量长时间基线链式配准方法。在“几何配准+增强谱分集配准”框架下,通过选定相邻已配准的影像作为第二主影像,与第一主影像共同形成配准干涉链。配准过程中,分别求取辅影像与第二主影像、采样后的辅影像与第一主影像的偏移量值,再通过偏移量叠加的方式求取最终的主辅影像配准偏移量值,从而达到TOPS模式的配准精度要求。实验表明,该方法可以有效避免长时间基线下配准精度差、相位跳变问题,其数据配准精度均优于0.001个像素。

InSAR is an important technical mean for the detection of the mined-out area,landslide and other geological disasters in the stage of railway survey,design and route selection.In order to solve the problems of phase jump and low registration accuracy of long-term baseline data under TOPS mode,a long-term baseline chain registration method considering the offset of adjacent images was proposed.Under the framework of“geometric registration and enhanced spectral diversity registration”,the adjacent registered image was selected as the second main image,and the registration interference chain was formed with the first main image.During the registration process,the offset values of the secondary image and the second main image,the sampled secondary image and the first main image were obtained respectively,and the final primary and secondary image registration offset values were obtained by the way of offset superposition,so as to meet the registration accuracy requirements of TOPS mode.The experimental results show that,the method can effectively avoid the problems of poor registration accuracy and phase jump under long baseline,and its data registration accuracy is better than 0.001 pixels.