基于Hausdorff距离的图像匹配并行算法设计与实现  

Design and Implementation of a Parallel Image Matching Algorithm Based on Hausdorff Distance

作　　者：于二丽[1] 周宁宁[1] 

机构地区：[1]南京邮电大学计算机学院,江苏南京210003

出　　处：《计算机技术与发展》2011年第9期28-31,共4页Computer Technology and Development

基　　金：软件开发环境国家重点实验室开放课题(BUAA-SKLSDE-09KF-03);国家重点基础研究发展计划(973)(2005CB321901)

摘　　要：随着图像匹配的应用越来越广泛,图像匹配的实时性要求也越来越高。为了提高图像匹配的速度和更好地利用多核计算资源,设计了一种基于Hausdorff距离的图像匹配并行算法。首先介绍了Hausdorff距离的定义,然后分析了图像匹配串行算法的效率,在此基础上设计了基于Hausdorff距离的图像匹配并行算法,最后采用Matlab在多核计算机上对并行算法进行了实现。实验结果表明,文中所设计的并行算法能够显著提高图像匹配速度,并具有较好的抗失真和抗噪声性能。文中设计的并行算法有较好的扩展性,可以将这种并行思想应用到其它图像匹配算法的并行设计中。With extensive application of image matching,real-time requirements of image matching are also increasing.In order to improve the speed of image matching and use the multi-core computing resources efficiently,a parallel image matching algorithm based on Hausdorff distance is designed.Firstly it introduces the definition of Hausdorff distance,and analyzes the efficiency of serial image matching algorithm based on Hausdorff distance.On this basis,a parallel image matching algorithm based on Hausdorff distance is designed and implemented in multi-core computer using Matlab.The experimental results show that the parallel algorithm is designed can greatly improve matching speed,and has good anti-distortion and anti-noise performance.This parallel idea presented in this paper has good scalability,and can be applied to other parallel image matching algorithms＇ design.

关 键 词：HAUSDORFF距离 图像匹配 多核计算机 并行算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]