基于最小二乘互相关算法的图像定位匹配研究

An image orientation algorithm based on least squares correlative matching method

提出了一种基于最小二乘互相关的图像定位匹配算法。该算法将图像互相关信息和最小二乘法结合实现图像定位匹配,匹配精度可以达到亚像素甚至1/100像素级,同时利用金字塔分层来提高定位匹配速度;通过本算法在印刷品质量自动化检测系统中的应用,验证了该算法的高精度与高速度特性。

An image orientation algorithm is presented in this paper, which is based on least squares correlative matching method.This algorithm combins image correlation information with least squares matching (LSM) to realize image orientation matching. By using pyramid algorithm, it also compensates the shortage of large computation caused by LSM. It has been proved in auto-detect system of actual printing quality that this algorithm has characteristics of high precision and high speed.