基于主成分分析的图像匹配方法研究

景像匹配和相关跟踪系统中,由于所面临的都是变化的场景,实时获取的图像与预存的基准图之间存在差异。传统的基于图像空间的匹配方法不能很好地克服这些差异给图像匹配定位带来的影响,使得完全按图像灰度特征的相关匹配应用受到限制。文中将主成分分析方法引入了图像匹配和相关跟踪过程中,提出一种能抗御一定图像畸变的基于主成分分析的图像匹配算法,并应用该方法进行了初步实验,证明该算法具有较高的匹配跟踪鲁棒性。

Research of an image matching method based on principal component analysis

In scene\|matching and correlative tracking systems, scenes change incessantly, so there is difference between real time images and reference images which are stored in advance. Traditional matching methods based on grey characteristics can not favorably cope with such difference brought about by scenic changes, limiting the applications of correlative matching based on grey characteristics. The method of putting principal component analysis into image matching and correlative tracking is introduced, and an image matching algorithm based on principal component analysis is proposed, which can bear some image distortions in image matching and tracking. Then the experimental results are presented to verify the robustness of this image matching method.