摘要
传统图象匹配算法利用相关法进行逐点计算,计算量大且做了许多无用功。该文提出一种利用图象灰度局部特征的约束,首先对图象进行粗匹配找出可能出现的区域,再进行精确匹配来定位,该方法计算量小,从而提高搜索效率。
Normalized pattern matching is based on correlation and operated point by point with a large quantum of operation and low efficiency.This paper gives a kind of restriction using local character of image grayscale.First a general matching will find the possible appearing area of image,then an accurate matching will locate it.The quantum of this kind of operation is very small,thus improves the efficiency of search..
出处
《计算机工程与应用》
CSCD
北大核心
2002年第9期101-102,共2页
Computer Engineering and Applications
