基于角点特征的SAR图像与光学图像匹配算法

The Matching Arithmetic of SAR Imagesand Optical Images Based on the Corner Features

摘要:

INS/SAR组合导航系统对SAR实时图像与光学参考图像的匹配算法提出了实时性和准确性要求。文章首先采用一种结合Harries算子与曲率尺度空间的算法,提取SAR图像与光学图像中的角点；然后利用距离约束、角点值约束、边缘约束进行角点匹配。经实验验证,该方法实时性好,匹配精度高,可较好的应用于INS/SAR组合导航系统的图像匹配中。

Abstract:

INS/SAR integrated navigation system required the matching method of SAR images and optical images to be fast and precise. In this paper, the corner features was extracted using a method combining curvature scale space and Harris operator firstly, and then the corner feature was matched using restriction of distance, coner value and edges. Being proved via the experiment, the main-line-pairs matching method in the article acted well and met the requirement of the navigation system.