Structure-Aware Progressive Multi-Modal Fusion Network for RGB-T Crack Segmentation
Journal of Imaging · 01.11.2025 · 11 (11) · 384
DOI: 10.3390/jimaging11110384 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Imaging, online erschienen am 01.11.2025. Verfasst von Zhengrong Yuan und 6 weiteren Autor:innen. Beteiligte Einrichtung: Hunan Architectural Design Institute Group Co., Ltd., Changsha 410208, China. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the Science and Technology Major Project of Changsha , the Research and Development Plan of Key Areas in Hunan Province, Science and Technology Major Project of Changsha, Research and Development Plan of Key Areas in Hunan Province.
Abstract (Originalsprache)
Crack segmentation in images plays a pivotal role in the monitoring of structural surfaces, serving as a fundamental technique for assessing structural integrity. However, existing methods that rely solely on RGB images exhibit high sensitivity to light conditions, which significantly restricts their adaptability in complex environmental scenarios. To address this, we propose a structure-aware progressive multi-modal fusion network (SPMFNet) for RGB-thermal (RGB-T) crack segmentation. The main idea is to integrate complementary information from RGB and thermal images and incorporate structural priors (edge information) to achieve accurate segmentation. Here, to better fuse multi-layer features from different modalities, a progressive multi-modal fusion strategy is designed. In the shallow encoder layers, two gate control attention (GCA) modules are introduced to dynamically regulate the fusion process through a gating mechanism, allowing the network to adaptively integrate modality-specific structural details based on the input. In the deeper layers, two attention feature fusion (AFF) modules are employed to enhance semantic consistency by leveraging both local and global attention, thereby facilitating the effective interaction and complementarity of high-level multi-modal features. In addition, edge prior information is introduced to encourage the predicted crack regions to preserve structural integrity, which is constrained by a joint loss of edge-guided loss, multi-scale focal loss, and adaptive fusion loss. Experimental results on publicly available RGB-T crack detection datasets demonstrate that the proposed method outperforms both classical and advanced approaches, verifying the effectiveness of the progressive fusion strategy and the utilization of the structural prior.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Hunan Architectural Design Institute Group Co., Ltd., Changsha 410208, China.
- School of Artificial Intelligence and Robotics, Hunan University, Changsha 410012, China.
- College of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, China.
Förderung
- the Science and Technology Major Project of Changsha , the Research and Development Plan of Key Areas in Hunan Province
- Science and Technology Major Project of Changsha
- Research and Development Plan of Key Areas in Hunan Province
Zitieren
APA
Yuan, Z., Ding, X., Xia, X., He, Y., Fang, H., Yang, B., & Fu, W. (2025). Structure-Aware Progressive Multi-Modal Fusion Network for RGB-T Crack Segmentation. Journal of Imaging, 11(11), 384. https://doi.org/10.3390/jimaging11110384
Vancouver
Yuan Z, Ding X, Xia X, He Y, Fang H, Yang B, et al. Structure-Aware Progressive Multi-Modal Fusion Network for RGB-T Crack Segmentation. Journal of Imaging. 2025;11(11):384. doi:10.3390/jimaging11110384
BibTeX
@article{Yuan2025,
author = {Yuan, Zhengrong and Ding, Xin and Xia, Xinhong and He, Yibin and Fang, Hui and Yang, Bo and Fu, Wei},
title = {Structure-Aware Progressive Multi-Modal Fusion Network for RGB-T Crack Segmentation},
journal = {Journal of Imaging},
year = {2025},
volume = {11},
number = {11},
pages = {384},
doi = {10.3390/jimaging11110384}
}
EUVIMED