基于对数极坐标的图像匹配综述  

Review of Image Matching Based on Logarithmic Polar Coordinates

作　　者：冉洪成 RAN Hong-cheng(College of Computer Science,Sichuan University,Chengdu 610065)

机构地区：[1]四川大学计算机学院,成都610065

出　　处：《现代计算机》2020年第4期65-69,共5页Modern Computer

摘　　要：首先介绍图像匹配的原理,然后通过分析笛卡尔坐标系和对数极坐标系的优缺点,引出基于对数极坐标的图像匹配算法,并介绍Fourier-Mellin变换图像匹配和基于对数极坐标变换的特征匹配两个主要算法,最后分析基于对数极坐标图像匹配算法的优势和不足,并对其做出展望。First introduces the principle of image matching, and then through the analysis of the advantages and disadvantages of cartesian coordinate system and logarithmic coordinate system, which leads to the image matching algorithm based on logarithmic coordinates, and introduced the Fourier-Mellin transform image matching and feature matching based on logarithmic coordinates transform are two main algorithms, finally analyzes the advantages and disadvantages of image matching algorithm based on logarithmic coordinates, and the outlook.

关 键 词：图像匹配 对数极坐标 FOURIER-MELLIN变换 特征匹配 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]