基于改进SIFT特征和极坐标的图像匹配定位研究  

Image Locating and Matching Based on Improved SIFT and Polar Coordinates

作　　者：雷波 Lei Bo(National Key Laboratory of Air-based Information Perception and Fusion,Luoyang,Henan 471023;Luoyang Institute of Electro-optical Equipment,AVIC,Luoyang,Henan 471023)

机构地区：[1]空基信息感知与融合全国重点实验室,河南洛阳471023 [2]中国航空工业集团公司洛阳电光设备研究所,河南洛阳471023

出　　处：《现代工程科技》2024年第13期85-88,共4页Modern Engineering Technology

摘　　要：为解决图像配准过程中因图像旋转及缩放导致匹配误差较大的问题,提出了一种基于对数极坐标和改进的SIFT相结合的匹配算法。首先,通过对数极坐标变换方法进行图像粗匹配,实现对待匹配图像旋转及尺度的估计;其次,在粗匹配的基础上,利用信息熵原理提取子块的特征描述符;最后,利用欧氏距离进行匹配度量,完成图像配准,利用图像变换模型实现两幅图像中相同区域目标的匹配定位。实验结果表明,该算法能够实现两幅图像的高精度匹配与目标定位。To solve the problem of large matching errors caused by image rotation and scaling during image registration,this article proposes a matching algorithm based on a combination of logarithmic polar coordinates and improved SIFT.Firstly,rough image matching is performed using the logarithmic polar coordinate transformation method to estimate the rotation and scale of the image to be matched;On the basis of rough matching,extract the feature descriptors of sub blocks using the principle of information entropy;Finally,using Euclidean distance for matching measurement,image registration is completed,and an image transformation model is used to achieve matching and localization of targets in the same region in two images.The experimental results show that this algorithm can achieve high-precision matching and target localization of two images.

关 键 词：极坐标 SIFT特征 图像匹配 目标定位 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TN92[自动化与计算机技术—计算机科学与技术]