HCFNet: A SAM2-Based Hierarchical Cross-Branch Frequency-Aware Network for Industrial Surface Defect Segmentation
Sensors · 20.07.2026 · 26 (14) · 4597
DOI: 10.3390/s26144597 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 20.07.2026. Verfasst von Jiwei Yu und 5 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Equipment Reliability, Shenyang University of Chemical Technology, Shenyang 110142, China. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Liaoning Provincial Science and Technology Plan Project.
Abstract (Originalsprache)
Foundation models such as the Segment Anything Model 2 (SAM2) have demonstrated strong performance in image segmentation; however, their application to industrial defect detection faces significant challenges due to the substantial domain gap between natural and industrial images, insufficient sensitivity to fine-grained high-frequency structures, and reliance on manual prompts. To address these issues, this study proposes a Hierarchical Cross-Branch Frequency-Aware Network (HCFNet) to adapt SAM2 for prompt-free industrial defect segmentation. First, a Gated Adapter is introduced into the frozen SAM2 encoder, enabling efficient cross-domain transfer without massive parameter retraining, thereby effectively preserving the pre-trained visual priors. Secondly, a Laplacian-enhanced Auxiliary Branch is designed to explicitly amplify high-frequency components, compensating for the inherent perception limitations of the Transformer backbone and significantly awakening the model’s sensitivity to subtle defects like micro-cracks. Finally, a Cross-branch Multi-scale Fusion Module is proposed to seamlessly align and integrate global semantic information with local structural details in a unified manner, resolving heterogeneous feature distribution conflicts. Extensive experiments on the MVTec AD and VisA datasets demonstrate that the proposed method consistently outperforms SAM2-based baselines in terms of mIoU and mDice. This study establishes an effective approach for leveraging foundation models in automated industrial inspection and is expected to drive advancements in precise defect perception technologies.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute of Equipment Reliability, Shenyang University of Chemical Technology, Shenyang 110142, China.
- Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China.
- Liaoning Key Laboratory of Intelligent Technology for Chemical Process Industry, Shenyang University of Chemical Technology, Shenyang 110142, China.
- The State Key Laboratory of Robotics and Intelligent Systems, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China.
ORCID-Profile: Kecheng Zhou, Ting Wang
Förderung
- Liaoning Provincial Science and Technology Plan Project
Zitieren
APA
Yu, J., Zhou, K., Wang, T., Gan, H., Wang, Y., & Gao, S. (2026). HCFNet: A SAM2-Based Hierarchical Cross-Branch Frequency-Aware Network for Industrial Surface Defect Segmentation. Sensors, 26(14), 4597. https://doi.org/10.3390/s26144597
Vancouver
Yu J, Zhou K, Wang T, Gan H, Wang Y, Gao S. HCFNet: A SAM2-Based Hierarchical Cross-Branch Frequency-Aware Network for Industrial Surface Defect Segmentation. Sensors. 2026;26(14):4597. doi:10.3390/s26144597
BibTeX
@article{Yu2026,
author = {Yu, Jiwei and Zhou, Kecheng and Wang, Ting and Gan, Hongxiao and Wang, Yu and Gao, Shuzhi},
title = {HCFNet: A SAM2-Based Hierarchical Cross-Branch Frequency-Aware Network for Industrial Surface Defect Segmentation},
journal = {Sensors},
year = {2026},
volume = {26},
number = {14},
pages = {4597},
doi = {10.3390/s26144597}
}
EUVIMED