介绍了用极坐标格式算法处理条带模式合成孔径雷达数据时距离弯曲的校正方法。距离弯曲会给图像带来两种形式的失真,一种是keystone形式的几何失真,另一种是空变的散焦。在大成像场景、低载频和近距离等情况下,这些失真会变得尤其严重。距离弯曲不仅影响图像的质量,而且限制了成像区的大小。文中改进了通常采用的图像域重采样的方法校正几何失真,通过对图像域的空变滤波去除散焦,这些校正方法适用于聚束模式中正侧视和斜视,而且计算量较小,对斜视数据的校正不会增加运算量,并且通过参数仿真和实测数据处理的实验结果验证了算法的有效

In this paper,we describe methods in spotlight mode SAR image performed via PFA which are used to solve range curvature introduced by the assumption of strictly planar wavefronts.Range curvature typically leads to geometry distortion in fine resolution SAR image,and an associated defocus of scatters in the scene may also be significant.For the geometry distortion,an image resampling operation is feasibility as a post-processing step to PFA to correct the distortion.The correction can also be achieved by a more efficient but approximate way based on the preceding one,and it will be presented in this paper.Then,a space-variant filter is applied to compensate the defocusing effect in image domain to follow the geometry distortion correction.Fortunately,these methods mentioned above are general so they are suitable for broadside as well as squinted modes,with no more computational burden for squint-mode image.This paper describes the basic concept and the implementation of the correction,as well as provides results demonstrating the feasibility and efficiency of these methods.

TN957.52

国家自然科学基金60502030，航空科学基金05D52027