基于边缘相似性的异源图像匹配  

Matching Multi-Sensor Images Based on Edge Similarity

作　　者：李壮[1] 朱宪伟[1] 

机构地区：[1]国防科技大学航天与材料工程学院

出　　处：《飞行器测控学报》2011年第2期37-41,共5页Journal of Spacecraft TT&C Technology

基　　金：863项目支持(2007AA12Z121)

摘　　要：异源图像匹配是视觉导航、多源图像融合分析的关键步骤之一。对于成像机理差别较大的异源图像,如SAR图像和可见光图像,采用传统的异源图像匹配算法难以得到满意结果。本文提出一种基于边缘相似性的异源图像匹配方法,首先分别提取待匹配图像的边缘特征点集;然后计算基准图的边缘距离场;最后基于边缘相似性模型,通过实时图边缘图和基准图边缘距离场计算边缘相似度,寻找相似度最大的变换参数即为最终匹配参数。采用SAR与可见光图对方法进行了测试,结果表明,这种方法能够快速可靠地实现异源图像匹配。Matching multi-sensor images is a key step in vision navigation and multi-sensor image fusion.However,the imaging mechanisms of SAR images and optical images are very different and it is difficult to match them using a conventional multi-sensor image matching algorithm.A new matching method based on edge similarity is proposed in this paper.With the method,edge maps are detected in reference image and real-time image separately.Then,the edge distance field of the reference image is calculated.Similarity measure is computed through real-time edge map and reference edge distance field based on edge similarity model.The max edge similarity corresponds to the matching result.SAR images and optical images are used to test this method.Experiment results show that the method matches multi-sensor images efficiently and reliably.

关 键 词：图像处理 图像匹配 边缘相似性 异源图像 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]