一种新的序列图像匹配定位算法

景象匹配是飞行器导航定位中广泛采用的一项技术，如何提高图像的匹配概率，始终是该项技术中一个值得研究的课题。针对传统的利用单幅实时图像进行匹配时，存在的实时景象图过小、匹配特征不足，从而导致匹配效率过低的问题。文中研究了一种基于归一化灰度相关序列图像匹配算法，利用飞行器飞越匹配区时所得到的序列实时图像，在序列中各单幅图像匹配的基础上，利用一定的规则来综合单幅图像的匹配结果，从而得到最终匹配结果。实验证明，文中方法不仅能显著提高匹配性能，而且有较大的实用性。

A new matching algorithm based on serial images

Image matching is widely used in the guidance of aircraft. It is a key problem to improve the matching probability. The real time image is so small that the matching probability is lower if you use the traditional matching way. To improve the matching probability, a new matching algorithm based on serial images and normalized cross correlation is proposed. The serial real time images are taken when the aircraft crosses the matching area. The result is obtained from the matching results of each image based on given rules. Experiment shows the algorithm not only upgrades the matching probability efficiently but also has right smart practicability.