一种基于边缘测度的加权Hausdorff景象匹配方法  

A Scene Matching Method with Weighted Hausdorff Distance Based on Edge Measure

作　　者：凌志刚[1] 潘泉[1] 张绍武[1] 梁彦[1] 李耀军[1] 

机构地区：[1]西北工业大学自动化学院,西安710072

出　　处：《宇航学报》2009年第4期1633-1638,共6页Journal of Astronautics

基　　金：国家自然科学基金(60634030);高等学校博士学科点专项科研基金(20060699032);航空科学基金(2006ZC53037,2007ZC53037)

摘　　要：针对经典Hausdorff距离匹配算法中仅考虑图像间边缘点的几何关系,而未充分考虑边缘点边缘测度显著性差异对Hausdorff有向距离贡献不同的特点,提出一种基于边缘测度的加权Hausdorff距离(EM-WHD)景象匹配方法。该方法集合结构张量和局部均值比方法构造出边缘测度图像,并根据测度图提取图像边缘;在图像匹配过程中,基于边缘测度的显著性不同,对每个边缘点采用不同权值的加权Hausdorff距离进行图像匹配定位。仿真实验结果证明了本文方法的有效性和计算快速性。According to the classical methods based on Hausdorff distance which only consider the geometric relationship among the edge feature points and don＇ t be taken into account the difference contribution of feature points to image matching, a simple but efficient real-time image matching aided navigation algorithm with weighted Hausdorff distance based on edge measure （EM-WHD） is proposed. Given a reference image and a sensed image, edge measure function is constructed based on the structure tensor method and ratio of average method. Furthermore, edge points are extracted. The weighted Hausdorff distance based on edge measure function and edge measure difference between matched edge points is carried out to match images. Simulation results show that this novel method is effective and can increase the computational speed.

关 键 词：结构张量 均值比 边缘测度 加权HAUSDORFF距离 

分 类 号：TN911.73[电子电信—通信与信息系统]