一种应用Partial Hausdorff距离的多尺度图像配准方法  

A Multiresolution Algorithm for Image Registration with Partial Hausdorff Distance

作　　者：王跃[1] 毛士艺[1] 陈炜[1] 

机构地区：[1]北京航空航天大学203教研室

出　　处：《航空学报》2004年第6期593-597,共5页Acta Aeronautica et Astronautica Sinica

摘　　要：基于图像边缘特征提取的PartialHausdorff距离的配准方法,可以有效抑制由于多传感器的成像特性差异、边缘特征提取等原因对配准的影响。本文对基于PartialHausdorff距离的图像配准方法作了改进,采用多尺度的配准方法,避免了PartialHausdorff距离准则引起的无值和多值性问题,而且显著减少了计算量,实现了PartialHausdorff距离准则对于图像配准的优点。Image registration algorithms based on edge extraction with Partial Hausdorff distance can efficiently reduce the influence made by different imaging properties of mutisensors, edge extraction procedure etc. In this work,the image registration algorithm with Partial Hausdorff distance is improved and the multiresolution method is adopted so as to avoid the problem of singularity and ambiguity introduced by Partial Hausdorff distance rule. Furthermore the computation quantity is greatly reduced, while various advantages of Partial Hausdorff distance rule can be thoroughly realized.

关 键 词：图象配准 PARTIAL HAUSDORFF距离 多尺度 

分 类 号：V243[航空宇航科学与技术—飞行器设计]