一种图像相关匹配算法的改进及实现  

Improvement and Implementation of Image Correlation Matching Algorithm

作　　者：宋刈非[1] 郑丽[1] 

机构地区：[1]东北电子技术研究所,辽宁锦州121000

出　　处：《光电技术应用》2013年第4期61-64,共4页Electro-Optic Technology Application

摘　　要：序贯相似性检测算法(SSDA)是针对传统模板匹配算法提出的一种高效的图像匹配算法,对SSDA算法进行改进,利用序列来代替固定阈值,采用隔点采样以及粗匹配和精匹配相结合,进一步提高了计算效率。并在此基础上,采用Visual C++6.0对该算法进行了编程实现。实验证明,该算法比传统的SSDA算法计算速度更快、匹配误差更小。A high efficient sequential similarity detection algorithm（SSDA） is proposed compared with the traditional mould board matching algorithm.SSDA is improved such as using sequence instead of fixed thresh old,dot sampling,coarse and fine matching combination so as to increase calculation speed further.Based on this,Visual C＋＋ 6.0 is used to implement the algorithm programming.Experimental results show that the algo rithm has faster calculation speed and less matching error comparing with traditional algorithm.

关 键 词：VISUALC++ SSDA 序贯相似性检测算法 图像匹配 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]