基于R-D算法的合成孔径声纳插值方法比较

Comparison of Interpolation Methods in Synthetic Aperture Sonar Based on R-D Method

合成孔径声纳(SAS)中由于基阵相对目标的运动导致回波空间产生了距离徙动,距离徙动的出现将会较大地影响最终声纳的成像质量,目前主要采用插值的方法进行矫正。该文基于SAS中的距离-多普勒(R-D)算法,讨论了距离徙动的成因及矫正思想,详细介绍了几种用于徙动矫正的经典插值方法,并给出了其在SAS成像中的应用表达式,对其性能进行了比较。最后对点目标进行了仿真成像,给出了矫正前后的目标图像,并对不同插值方法的成像质量及运算量进行了分析比较。分析结果表明,这几种插值方法都改善了成像的质量,但拉格朗日法和香农法在改善图像方位向性能上更有优势,可在实际中应用。

The movement of the array in Synthetic aperture sonar （SAS） relative to the object causes range migration in echo space, which would affect the ultimate image and is mainly corrected by interpolation at present. Based on range - Doppler algorithm, this paper first discusses the cause and correction idea of range migration, gives several classic interpolation methods, then presents their detailed expressions used in SAS, followed by a comparison of their performance. Finally, the simulation of point targets is carried out, target images before and after correction are displayed, added with some comparison of imaging quality and computation load. It is demonstrated that all these methods can improve the image quality, but Lagrange method and Shannon method show an advantage in mending the azimuth direction, which can be used in practice.