一种改进的基于SURF的快速图像匹配算法研究  

An Improved Algorithm of Image Registration based on SURF

作　　者：卢选民[1] 孙志军[2] 吴健[2] 王君本[1] 俞天秀[2] 赵良[2] 丁晓宏[2] 

机构地区：[1]西北工业大学电子信息学院,陕西西安710072 [2]敦煌研究院数字中心,甘肃敦煌736200

出　　处：《敦煌研究》2010年第6期88-92,共5页Dunhuang Research

基　　金：国家文物局文化遗产保护科学和技术研究课题"敦煌壁画数字图像智能拼接系统"(2009140-22/05)

摘　　要：针对传统的图像特征匹配算法数据量大、计算耗时长的缺点,本文提出了一种改进的基于SURF的快速图像配准算法。该算法在积分图像的基础上进行快速计算,通过快速Hessian检测子来检测特征点。对于每个特征点,通过计算哈尔小波变换来确定其主方向,并确定特征点描述子,再根据描述向量之间的欧式距离实现图像间的特征点的匹配。实验表明,该算法既能满足匹配准确性的要求,又具有计算量小、计算速度快的优点。With the shortcomings of the large data amount and long time consuming in the conventional image feature matching algorithms,a new algorithm based on SURF for image registration is presented in this paper.It is fast computed based on the integral image and through the Fast-Hessian detector,the feature points are extracted.For each feature point,the dominant orientation is assigned by computing Haar-wavelet responses,and then the descriptor is generated.Image matching is made based on Euclidean distance between these descriptors.Experimental result shows that this algorithm can not only meet the requirement of accuracy,but also has a small data amount and fast speed for image registration.

关 键 词：图像匹配 SURF 积分图像 HESSIAN矩阵 尺度空间 特征点描述子 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]