基于主成分分析的灰度图像匹配方法研究  

Gray Image Matching Method Based on Principal Component Analysis

作　　者：熊文真[1] 陈秀琴[1] 李红娟[2] 

机构地区：[1]信阳职业技术学院,河南信阳464000 [2]昆明理工大学冶金节能减排教育部工程研究中心,昆明650093

出　　处：《襄阳职业技术学院学报》2016年第5期25-27,30,共4页Journal of Xiangyang Polytechnic

基　　金：昆明理工大学人才科研启动(141180901);NSFC-云南联合基金(U0937604)

摘　　要：为了解决灰度图像匹配方法中的归一化积相关算法在图像匹配过程中计算量大、各种噪声的干扰、匹配结果不准确等问题,文章提出了基于主成分分析的归一化积相关算法,即利用图像灰度值的主成分构造一阶微分方程进行匹配,这样就把整幅图像的灰度值降低为较少的几个主要变量后再进行匹配,从而减少了计算量,一定程度上排除了各种噪声的干扰,使两幅图像快速进行匹配,实验结果证明该算法的有效性。In order to solve the problems of normalized product algorithm in image matching process such as large amount of calculation, all kinds of noise interference, inaccurate matching results etc., the paper proposed the normalized product algorithm based on principal component analysis, namely using the image grey value matching principal component structure of first order differential equation, thus lowered the whole image grey value for less match again after several main variables, which reduced the amount of calculation as well as various types of noise interference, making the two images match quickly. The experiment results show the effectiveness of the proposed algorithm.

关 键 词：图像匹配 归一化积相关算法 主成分分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]