测绘科学技术学报2025，Vol.41Issue(4)：411-418,8.DOI:10.3969/j.issn.1673-6338.2025.04.011
基于结构特征增强的弱纹理SAR-光学图像匹配算法
Weak Texture SAR-optical Image Registration Algorithm Based on Structural Feature Enhancement
王凯 1党涛 2耿彦龙 2刘合凤2
作者信息
- 1. 西安测绘总站,陕西西安 710054||智慧地球重点实验室,陕西西安 710054
- 2. 西安测绘总站,陕西西安 710054
- 折叠
摘要
Abstract
Traditional image matching algorithms have practical limitations in addressing the task of matching weak-textured SAR-optical images,such as a low number of correct matches,poor algorithm robustness,and low matc-hing accuracy.This paper proposes a weak-textured SAR-optical image matching algorithm based on the idea of en-hancing image structural features.This method first performs local normalization filtering and histogram equalization enhancement on SAR and optical images.Then,it uses FAST and gray centroid method for feature extraction on the enhanced images.The extracted feature points are projected onto multi-scale images to endow the algorithm with scale invariance,and an improved HOG method is utilized for feature description.Finally,the algorithm combines nearest neighbor feature matching with FSC mismatch elimination to obtain precise matching results.Multiple sets of experimental results demonstrate that the proposed algorithm outperforms state-of-the-art methods in terms of both the number of correct matches and root mean square error.It effectively overcomes issues caused by nonlinear radiometric differences and weak textures,exhibits good rotation and scale invariance,and is capable of adapting to heterogeneous image matching tasks in complex scenarios.关键词
图像匹配/结构特征增强/弱纹理/特征匹配/模态差异Key words
image matching/structural feature enhancement/weak texture/feature matching/modal differences分类
测绘与仪器引用本文复制引用
王凯,党涛,耿彦龙,刘合凤..基于结构特征增强的弱纹理SAR-光学图像匹配算法[J].测绘科学技术学报,2025,41(4):411-418,8.
