摘要
图像匹配算法是计算机视觉应用研究的基础。为了解决传统尺度不变特征变换算法(Scale-invariant Feature Transform, SIFT)运行效率低、存在误匹配点对和精匹配点对稀少等问题,本文将其与快速视网膜关键点(Fast Retina Keypoint, FREAK)算法和PROSAC算法相结合,提出了一种SIFT-FREAK图像匹配算法。首先在特征点检测阶段,用SIFT算法提取具有尺度不变性的特征点,然后利用FREAK算法构建二进制描述子,特征匹配时,先采用汉明距离(Hamming Distance)进行初始匹配点对,再用双向匹配完成粗匹配,最后用PROSAC算法进行精匹配。实验结果表明,本文SIFT-FREAK算法比SIFT算法和FREAK算法(以加速分割检测特征(Feature from Accelerated Segment Test, FAST)算法提取特征点)在准确率、运行效率和精匹配点数三大方面都有着一定的优势。
Image matching algorithm is the basis of computer vision application research.In order to solve the problems of low operating efficiency of traditional scale-invariant feature transform(SIFT),mismatched point pairs and rare precision matching point pairs.This paper combines it with the Fast Retina Keypoint(FREAK)algorithm and the PROSAC algorithm,and proposes a SIFT-FREAK image matching method.First,in the feature point detection stage,the SIFT algorithm is used to extract the feature points with scale invariance,and then the FREAK algorithm is used to construct a binary descriptor.When the feature is matched,the Hamming Distance is used for the initial matching point pair,and then the two-way matching is used for rough matching,and the PROSAC algorithm is used for fine matching.The overall results of the experiment show that the SIFT-FREAKE algorithm in this paper is more accurate than the SURF algorithm and the FREAK algorithm(Feature from Accelerated Segment Test(FAST)algorithm to extract feature points)in accuracy,operating efficiency and precision matching points.There are obvious advantages in all aspects.
作者
陈建新
唐丽玉
CHEN Jianxin;TANG Liyu(Key Laboratory of Spatial Data Mining&Information Sharing of Ministry of Education,Fuzhou University,Fuzhou 350108,China;National Engineering Research Center of Geospatial Information Technology,Fuzhou University,Fuzhou 350108,China)
出处
《测绘与空间地理信息》
2023年第6期32-35,共4页
Geomatics & Spatial Information Technology
基金
国家重点研发计划项目(2018YFB1004905)资助。
作者简介
陈建新(1996-),男,福建福州人,测绘工程专业硕士研究生,主要研究方向为计算机视觉和增强现实。;通信作者:唐丽玉(1972-),女,福建莆田人,研究员,博士,2007年毕业于福州大学通信与信息系统专业,主要从事地学可视化与虚拟地理环境、虚拟植物等方面的教学与科研工作。
