一种基于SIFT-NMI的多源影像配准方法  

A Multi-source Image Registration Method Based on SIFT-NMI

作　　者：徐绍哲 詹总谦[1] XU Shaozhe;ZHAN Zongqian(School of Geodsy and Geomatics,Wuhan University,Wuhan 430079,China)

机构地区：[1]武汉大学测绘学院,湖北武汉430079

出　　处：《测绘地理信息》2022年第6期32-35,共4页Journal of Geomatics

基　　金：国家自然科学基金(61871295)。

摘　　要：针对不同尺度的无人机多源遥感影像,在使用尺度不变特征变换(scale invariant feature transform,SIFT)粗配准的基础上,加入归一化互信息(normalized mutual information,NMI)进行精配准,有效克服了多源影像的尺度和光谱差异性问题,实现了无人机多源影像的自动配准。实验结果表明,所提方法鲁棒性强,是一种有效的多源影像配准方法。Based on the coarse registration by SIFT,normalized mutual information is added for precise registration.Thus the problem that the difference of scale and spectrum in multisource image registration is effectively resolved,and then the automatic registration of UAV multi-source images is realized.The experimental result shows that this method is an effective multi-source image registration method which has strong robustness.

关 键 词：尺度不变特征变换 随机抽样一致算法 归一化互信息 无人机影像 多源影像配准 

分 类 号：P237[天文地球—摄影测量与遥感]