基于Fourier-Mellin的图像配准算法研究  

Study on an Image Registration Algorithm Based on Fourier-Mellin

作　　者：郭清宇[1] 崔伟民[1] 

机构地区：[1]中原工学院,郑州450007

出　　处：《中原工学院学报》2012年第6期26-29,共4页Journal of Zhongyuan University of Technology

基　　金：河南省科技攻关计划项目(0624220100)

摘　　要：提出了一种基于Fourier-Mellin的图像配准算法.通过对图像进行加窗、滤波等处理,消除了图像的边界效应产生的伪高频部分,减少了混叠现象,增大了图像的信噪比,提高了图像相关算法的有效性和精度.实验结果表明本算法有效.This paper proposes an image registration algorithm based on Fourier-Mellin. The images using some techniques such as windowing and filtering, which removes spurious high-frequencies in spectral leakage created by the image boundary, greatly reduces the ratationally dependent aliasing, increases the signal- to-noise ratio of the images and improves the efficiency and accuracy of the phase correlation algorithm. Experimental results demonstrate the efficiency of the proposed algorithm.

关 键 词：图像配准 FOURIER-MELLIN变换 尺度不变性 离散傅里叶变换 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]