论文部分内容阅读
采用地图与参考模板匹配的方法找出合成孔径雷达图象中的预选特征能提供导航定位数据。这里提出的图象匹配法是把参考特征编码到二进制参考模板中,利用这个参考模板提供自适应规范,分割雷达图象,然后进行简单的二进制相关。这个方法与传统的互相关处理相比,大大简化了计算的复杂性,并且在竞争的特征出现时,也能以很高的概率去获得参考特征。本文在理论上分析了这种图象匹配方法的性能特征,给出了利用合成孔径雷达图象匹配方法得到的几例结果。
The method of map matching with reference template is used to find the preselected features in synthetic aperture radar images to provide navigation and positioning data. The image matching method proposed here encodes the reference features into a binary reference template, which provides an adaptive specification, splits the radar image, and then performs a simple binary correlation. Compared with the traditional cross-correlation processing, this method greatly simplifies the computational complexity and obtains the reference features with high probability when competing features appear. In this paper, the performance characteristics of this image matching method are analyzed theoretically, and several results obtained by using the synthetic aperture radar image matching method are given.