SGCM: Semantic and Geometric Consistency for Robust Aerial Image Matching

Aerial image matching remains an ongoing challenge due to the significant geometric distortions and nonlinear radiation distortion. Most current approaches primarily focus on low-level image features, while overlooking the potential of high-level semantic information to guide the matching process, l...

Full description

Saved in:
Bibliographic Details
Main Authors: Xiangzeng Liu, Guanglu Shi, Chi Wang, Xiaodong Zhang, Qiguang Miao
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11105454/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849404354154463232
author Xiangzeng Liu
Guanglu Shi
Chi Wang
Xiaodong Zhang
Qiguang Miao
author_facet Xiangzeng Liu
Guanglu Shi
Chi Wang
Xiaodong Zhang
Qiguang Miao
author_sort Xiangzeng Liu
collection DOAJ
description Aerial image matching remains an ongoing challenge due to the significant geometric distortions and nonlinear radiation distortion. Most current approaches primarily focus on low-level image features, while overlooking the potential of high-level semantic information to guide the matching process, limiting the robustness and accuracy of these methods in complex scenarios. In this article, we propose a novel coarse-to-fine aerial image matching framework called SGCM, which considers both semantic and geometric consistency in feature representation. Concretely, in the coarse matching stage, the SAM foundation model is leveraged to efficiently extract salient structural regions, providing valuable semantic guidance for subsequently matching procedure. To achieve accurate region matching, we construct the region description graphs that incorporate both semantic categories and geometric attributes information. A graph neural network (GNN) is then employed for feature aggregation and node descriptor updates, leading to more robust region representations. Coarse matching is performed by computing a similarity matrix between node descriptors, which helps effectively mitigate rotational and scaling differences between input images. Consequently, a fine matching strategy combining intraregion point matching and global matching is developed, which significantly improves the robustness of image matching under geometric variations. The performance of our method has been thoroughly evaluated on three challenging aerial image datasets and compared with five state-of-the-art matching methods. Extensive experimental results demonstrate the substantial advantages of SGCM in terms of matching accuracy and efficiency.
format Article
id doaj-art-b3034c5148c74c13a9e8366fa0faf7db
institution Kabale University
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-b3034c5148c74c13a9e8366fa0faf7db2025-08-20T03:37:01ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-0118196961970910.1109/JSTARS.2025.359412711105454SGCM: Semantic and Geometric Consistency for Robust Aerial Image MatchingXiangzeng Liu0https://orcid.org/0000-0002-2751-6096Guanglu Shi1https://orcid.org/0009-0008-7185-4763Chi Wang2https://orcid.org/0009-0006-3230-1483Xiaodong Zhang3Qiguang Miao4https://orcid.org/0000-0001-6766-8310School of Computer Science and Technology, Xidian University, Xi’an, ChinaSchool of Computer Science and Technology, Xidian University, Xi’an, ChinaSchool of Computer Science and Technology, Xidian University, Xi’an, ChinaSchool of Computer Science and Technology, Xidian University, Xi’an, ChinaSchool of Computer Science and Technology, Xidian University, Xi’an, ChinaAerial image matching remains an ongoing challenge due to the significant geometric distortions and nonlinear radiation distortion. Most current approaches primarily focus on low-level image features, while overlooking the potential of high-level semantic information to guide the matching process, limiting the robustness and accuracy of these methods in complex scenarios. In this article, we propose a novel coarse-to-fine aerial image matching framework called SGCM, which considers both semantic and geometric consistency in feature representation. Concretely, in the coarse matching stage, the SAM foundation model is leveraged to efficiently extract salient structural regions, providing valuable semantic guidance for subsequently matching procedure. To achieve accurate region matching, we construct the region description graphs that incorporate both semantic categories and geometric attributes information. A graph neural network (GNN) is then employed for feature aggregation and node descriptor updates, leading to more robust region representations. Coarse matching is performed by computing a similarity matrix between node descriptors, which helps effectively mitigate rotational and scaling differences between input images. Consequently, a fine matching strategy combining intraregion point matching and global matching is developed, which significantly improves the robustness of image matching under geometric variations. The performance of our method has been thoroughly evaluated on three challenging aerial image datasets and compared with five state-of-the-art matching methods. Extensive experimental results demonstrate the substantial advantages of SGCM in terms of matching accuracy and efficiency.https://ieeexplore.ieee.org/document/11105454/Aerial image matchinggraph neural network (GNN)SAM modelsemantic and geometric consistency
spellingShingle Xiangzeng Liu
Guanglu Shi
Chi Wang
Xiaodong Zhang
Qiguang Miao
SGCM: Semantic and Geometric Consistency for Robust Aerial Image Matching
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Aerial image matching
graph neural network (GNN)
SAM model
semantic and geometric consistency
title SGCM: Semantic and Geometric Consistency for Robust Aerial Image Matching
title_full SGCM: Semantic and Geometric Consistency for Robust Aerial Image Matching
title_fullStr SGCM: Semantic and Geometric Consistency for Robust Aerial Image Matching
title_full_unstemmed SGCM: Semantic and Geometric Consistency for Robust Aerial Image Matching
title_short SGCM: Semantic and Geometric Consistency for Robust Aerial Image Matching
title_sort sgcm semantic and geometric consistency for robust aerial image matching
topic Aerial image matching
graph neural network (GNN)
SAM model
semantic and geometric consistency
url https://ieeexplore.ieee.org/document/11105454/
work_keys_str_mv AT xiangzengliu sgcmsemanticandgeometricconsistencyforrobustaerialimagematching
AT guanglushi sgcmsemanticandgeometricconsistencyforrobustaerialimagematching
AT chiwang sgcmsemanticandgeometricconsistencyforrobustaerialimagematching
AT xiaodongzhang sgcmsemanticandgeometricconsistencyforrobustaerialimagematching
AT qiguangmiao sgcmsemanticandgeometricconsistencyforrobustaerialimagematching