{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T12:15:21Z","timestamp":1790079321921,"version":"4.0.1"},"reference-count":36,"publisher":"MDPI AG","issue":"9","license":[{"start":{"date-parts":[[2026,9,20]],"date-time":"2026-09-20T00:00:00Z","timestamp":1789862400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Biomimetics"],"abstract":"<jats:p>Automated vision-based structural health monitoring (SHM) of catenary support components is critical for ensuring railway operational safety. However, achieving reliable structural damage identification in practical engineering scenarios is hindered by complex environmental and operational variations (EOVs) in aerial imagery, the multi-scale nature of structural degradation, and the scarcity of damage samples. To address these SHM challenges, this paper proposes a high-precision unsupervised damage detection framework named Mask-Guided Asymmetrical Flow (MGAF). First, to mitigate the impact of EOVs, a SAM-guided preprocessing strategy is introduced to explicitly suppress background interference and extract effective structural Regions of Interest (ROIs). Second, an Asymmetrical Parallel Flow architecture is designed to balance damage detection sensitivity and processing latency. By optimizing the flow depth for high-resolution features, this architecture prevents overfitting to high-frequency environmental noise while preserving deep semantic information. Furthermore, a Cross-Scale Feature Fusion (CSFF) module is developed to enhance the detection capability for early-stage structural damages (e.g., fatigue micro-cracks and fastener looseness) by integrating global structural semantics with local textural damage indicators. Finally, a Morphological Edge Suppression (MES) mechanism is employed to eliminate boundary artifacts, thereby reducing the false alarm rate in pixel-level damage localization. Extensive experiments on the self-constructed CSCUD dataset demonstrate that the proposed system achieves competitive anomaly detection performance compared with representative unsupervised methods with an Image-level AUROC of 98.9% and a Pixel-level AP of 40.1%. During online inference, MGAF achieves a processing time of 0.077 s per frame on an NVIDIA RTX 4090 GPU, excluding the offline GSI-Net localization and SAM-based structural ROI extraction stages. This demonstrates the efficiency of the proposed anomaly detection module for practical railway inspection scenarios. Additional experiments on selected categories from the MVTec AD benchmark provide preliminary evidence of the transferability of MGAF to visually similar industrial anomaly detection scenarios.<\/jats:p>","DOI":"10.3390\/biomimetics11090679","type":"journal-article","created":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T11:09:35Z","timestamp":1790075375000},"page":"679","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Vision-Based Structural Health Monitoring of Catenary Components in UAV Inspections via Mask-Guided Asymmetrical Flow"],"prefix":"10.3390","volume":"11","author":[{"given":"Qiaolu","family":"Wang","sequence":"first","affiliation":[{"name":"School of Artificial Intelligence and Electronic Engineering, Sichuan Technology and Business University, No. 65, Xueyuan Road, Pidu District, Chengdu 611745, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"software","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"},{"role":"formal-analysis","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"},{"role":"writing-original-draft","vocabulary":"credit"},{"role":"investigation","vocabulary":"credit"},{"role":"data-curation","vocabulary":"credit"}]},{"given":"Haitao","family":"Lan","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence and Electronic Engineering, Sichuan Technology and Business University, No. 65, Xueyuan Road, Pidu District, Chengdu 611745, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"visualization","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"},{"role":"formal-analysis","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"},{"role":"software","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"project-administration","vocabulary":"credit"}]},{"given":"Tianyu","family":"Zhou","sequence":"additional","affiliation":[{"name":"School of Electrical Engineering, Southwest Jiaotong University, Chengdu 610097, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"investigation","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"data-curation","vocabulary":"credit"},{"role":"software","vocabulary":"credit"},{"role":"funding-acquisition","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"},{"role":"project-administration","vocabulary":"credit"}]},{"given":"Ning","family":"Ma","sequence":"additional","affiliation":[{"name":"School of Electrical Engineering, Southwest Jiaotong University, Chengdu 610097, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"investigation","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"data-curation","vocabulary":"credit"},{"role":"writing-original-draft","vocabulary":"credit"},{"role":"funding-acquisition","vocabulary":"credit"}]},{"given":"Haonan","family":"Yang","sequence":"additional","affiliation":[{"name":"School of Electrical Engineering, Southwest Jiaotong University, Chengdu 610097, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"supervision","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"investigation","vocabulary":"credit"},{"role":"data-curation","vocabulary":"credit"},{"role":"project-administration","vocabulary":"credit"},{"role":"funding-acquisition","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8007-4899","authenticated-orcid":false,"given":"Jingke","family":"Yan","sequence":"additional","affiliation":[{"name":"State Key Laboratory of Rail Transit Vehicle System, Southwest Jiaotong University, Chengdu 610000, China"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"investigation","vocabulary":"credit"},{"role":"visualization","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"project-administration","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]}],"member":"1968","published-online":{"date-parts":[[2026,9,20]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"6319","DOI":"10.1109\/TITS.2026.3680498","article-title":"Graph-MDETR: A graph-guided Mamba-DETR network for UAV catenary support components detection in electrified railways","volume":"27","author":"Yang","year":"2026","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"14100","DOI":"10.1109\/TTE.2025.3609347","article-title":"Generalized Koopman neural operator for data-driven modelling of electric railway pantograph-catenary systems","volume":"11","author":"Wang","year":"2025","journal-title":"IEEE Trans. Transp. Electrif."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"116170","DOI":"10.1016\/j.engappai.2026.116170","article-title":"Hybrid Spiking-Visual-Language Framework for Unmanned Aerial Vehicle-Based Inspection and Anomaly Detection of Catenary Support Components in Electrified Railways","volume":"182","author":"Ma","year":"2026","journal-title":"Eng. Appl. Artif. Intell."},{"key":"ref_4","first-page":"1370","article-title":"Fault detection of catenary insulator damage\/foreign material based on corner matching and spectral clustering","volume":"35","author":"Zhang","year":"2014","journal-title":"Chin. J. Sci. Instrum."},{"key":"ref_5","first-page":"30","article-title":"Foreign body detection between insulator pieces in electrified railway based on affine moment invariant","volume":"35","author":"Yang","year":"2013","journal-title":"J. China Railw. Soc."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"839","DOI":"10.1007\/s10845-015-1136-3","article-title":"A new arc detection method based on fuzzy logic using S-transform for pantograph\u2013catenary systems","volume":"29","author":"Karakose","year":"2018","journal-title":"J. Intell. Manuf."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"1109","DOI":"10.1109\/TITS.2020.3020287","article-title":"Adversarial reconstruction based on tighter oriented localization for catenary insulator defect detection in high-speed railways","volume":"23","author":"Zhong","year":"2020","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_8","first-page":"5024214","article-title":"ECF-STPM: A robust crack detection method for railway catenary components","volume":"72","author":"Zhang","year":"2023","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"257","DOI":"10.1109\/TIM.2017.2775345","article-title":"Automatic defect detection of fasteners on the catenary support device using deep convolutional neural network","volume":"67","author":"Chen","year":"2017","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"5020413","DOI":"10.1109\/TIM.2024.3398091","article-title":"MTA-Net: A one-stage detector based on a multiscale task-aligned network for catenary support components","volume":"73","author":"Xie","year":"2024","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"742","DOI":"10.1177\/14759217241290537","article-title":"Enhancing fault diagnosis with a hybrid attention mechanism and spatio-temporal feature mining model using small sample data","volume":"25","author":"He","year":"2026","journal-title":"Struct. Health Monit."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"350","DOI":"10.1080\/01969722.2025.2606002","article-title":"VinsFusion-Line: Binocular vision inertial navigation real-time SLAM system based on line features","volume":"57","author":"Yang","year":"2026","journal-title":"Cybern. Syst."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"13637","DOI":"10.1109\/TCSVT.2026.3690701","article-title":"WaveletMixer: When wavelet transform meets self-attention in single image super-resolution","volume":"36","author":"Zhang","year":"2026","journal-title":"IEEE Trans. Circuits Syst. Video Technol."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Chen, Z., Li, N., Zhang, H., Xu, S., Chen, W., Zhang, J., Fu, L., Hu, H., Peng, Y., and Deng, F. (2025). Direction-aware attention for detail-preserving terahertz image super-resolution. Nondestruct. Test. Eval., 1\u201324.","DOI":"10.1080\/10589759.2025.2584623"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"797","DOI":"10.1177\/14759217241298490","article-title":"An enhanced deep intelligent model with feature fusion and ensemble learning for the fault diagnosis of rotating machinery","volume":"25","author":"Zhuang","year":"2026","journal-title":"Struct. Health Monit."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"951","DOI":"10.1177\/14759217241291268","article-title":"Prototype-assisted multiscale graph representation learning-based mechanical fault detection method under complex operating conditions","volume":"25","author":"Xiang","year":"2026","journal-title":"Struct. Health Monit."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"1340","DOI":"10.1177\/14759217241307575","article-title":"A geometrical morphology-enhanced computer vision approach for structural health assessment","volume":"25","author":"Li","year":"2026","journal-title":"Struct. Health Monit."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"765","DOI":"10.1080\/01969722.2025.2573330","article-title":"Enhancing medical diagnosis through multimodal image fusion: A novel approach using modified swin-based cross attention fusion","volume":"57","author":"Sreekala","year":"2026","journal-title":"Cybern. Syst."},{"key":"ref_19","first-page":"1961","article-title":"Illustration of image registration-based novel segmentation and classification model using multimodal medical images with 3d-trrsegnet and adaptive ran","volume":"57","author":"Nelakuditi","year":"2025","journal-title":"Cybern. Syst."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"37","DOI":"10.1177\/14759217241270919","article-title":"Enhancing automated health monitoring of road infrastructure through a hierarchical robust deep learning approach","volume":"25","author":"Deng","year":"2026","journal-title":"Struct. Health Monit."},{"key":"ref_21","first-page":"1996","article-title":"Enhanced transfer learning-based CNN for abnormal human activity detection in video surveillance using spatial-temporal features","volume":"57","author":"Kumar","year":"2025","journal-title":"Cybern. Syst."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"866","DOI":"10.1080\/01969722.2025.2590761","article-title":"Mobile-Le Harmonic Fusion Network for Object Recognition and SiamMoT Based Multi-Object Tracking Using Video Surveillance","volume":"57","author":"Shajeena","year":"2026","journal-title":"Cybern. Syst."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Chen, Z., Zhang, H., Fan, M., and Li, N. (2026). Wavelet-Guided Dynamic Transformer for High-Fidelity Super-Resolution in Terahertz Imaging Systems. IEEE Sens. J.","DOI":"10.1109\/JSEN.2026.3730211"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"111563","DOI":"10.1016\/j.knosys.2024.111563","article-title":"REB: Reducing biases in representation for industrial anomaly detection","volume":"290","author":"Lyu","year":"2024","journal-title":"Knowl.-Based Syst."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"4571","DOI":"10.52202\/068431-0330","article-title":"A unified model for multi-class anomaly detection","volume":"35","author":"You","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"71162","DOI":"10.52202\/079017-2274","article-title":"Mambaad: Exploring state space models for multi-class unsupervised anomaly detection","volume":"37","author":"He","year":"2024","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Yan, J., Chen, B., Zhang, F., Cheng, Y., Wang, H., Wang, H., Wang, M., Li, T., and Zhang, W. (2026). Meta-Learning-Based Graph Convolutional Wavelet Network for Intelligent Dynamic Modeling of High-Speed Rail Subsystems. IEEE Trans. Veh. Technol.","DOI":"10.1109\/TVT.2026.3714895"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"2510918","DOI":"10.1109\/TIM.2026.3687363","article-title":"Support structure detection of railway catenary systems on UAVs using spike-based brain-inspired neural network","volume":"75","author":"Ma","year":"2026","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"4413814","DOI":"10.1109\/TGRS.2024.3460085","article-title":"SAM-RSIS: Progressively adapting SAM with box prompting to remote sensing image instance segmentation","volume":"62","author":"Luo","year":"2024","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Deng, H., and Li, X. (2022). Anomaly detection via reverse distillation from one-class embedding. Proceedings of the 2022 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, LA, USA, 18\u201324 June 2022, IEEE.","DOI":"10.1109\/CVPR52688.2022.00951"},{"key":"ref_31","first-page":"8472","article-title":"A diffusion-based framework for multi-class anomaly detection","volume":"38","author":"He","year":"2024","journal-title":"Proc. AAAI Conf. Artif. Intell."},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Liu, Z., Zhou, Y., Xu, Y., and Wang, Z. (2023). Simplenet: A simple network for image anomaly detection and localization. Proceedings of the 2023 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada, 17\u201324 June 2023, IEEE.","DOI":"10.1109\/CVPR52729.2023.01954"},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Zhang, X., Li, S., Li, X., Huang, P., Shan, J., and Chen, T. (2023). Destseg: Segmentation guided denoising student-teacher for anomaly detection. Proceedings of the 2023 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada, 17\u201324 June 2023, IEEE.","DOI":"10.1109\/CVPR52729.2023.00381"},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Guo, J., Lu, S., Zhang, W., Chen, F., Li, H., and Liao, H. (2025). Dinomaly: The less is more philosophy in multi-class unsupervised anomaly detection. Proceedings of the 2025 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 10\u201317 June 2025, IEEE.","DOI":"10.1109\/CVPR52734.2025.01900"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"110828","DOI":"10.1016\/j.apacoust.2025.110828","article-title":"SDMAE: A self-supervised learning method based on a self-distillation masked autoencoder for anomalous sound detection","volume":"239","author":"Ye","year":"2025","journal-title":"Appl. Acoust."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"7140","DOI":"10.1109\/TPAMI.2025.3570494","article-title":"DiffusionAD: Norm-guided one-step denoising diffusion for anomaly detection","volume":"47","author":"Zhang","year":"2025","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."}],"container-title":["Biomimetics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2313-7673\/11\/9\/679\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T11:51:27Z","timestamp":1790077887000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2313-7673\/11\/9\/679"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,20]]},"references-count":36,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,9]]}},"alternative-id":["biomimetics11090679"],"URL":"https:\/\/doi.org\/10.3390\/biomimetics11090679","relation":{},"ISSN":["2313-7673"],"issn-type":[{"value":"2313-7673","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,20]]}}}