遮挡人脸识别技术综述
A Review of Occluded Face Recognition Techniques
DOI: 10.12677/csa.2025.155137, PDF,    科研立项经费支持
作者: 王晓彤, 韩金颖*, 王俊兴, 李振乾：北华航天工业学院计算机学院，河北 廊坊
关键词: 遮挡人脸识别；深度学习；特征提取；遮挡修复；Occlusion Face Recognition； Depth Learning； Feature Extraction； Occlusion Repair
摘要: 人脸识别作为一种非接触、便捷的生物识别技术，在安防、金融、身份认证等领域得到广泛应用。然而，面部遮挡会导致关键特征丢失，严重影响识别性能，成为该领域的重要挑战。针对这一问题，当前研究主要从两个方向展开：一是改进人脸识别算法本身，使其对遮挡具有更强的鲁棒性；二是通过遮挡修复技术还原完整人脸，再执行识别。本文系统综述了遮挡人脸识别技术的最新进展，首先分析直接识别方法，接着探讨基于遮挡修复的方法，并对比两种策略的优缺点及适用场景。最后，本文展望了未来研究方向，为遮挡人脸识别技术的进一步发展提供参考。
Abstract: As a non-contact and convenient biometric technology, face recognition is widely used in security, finance, identity authentication and other fields. However, facial occlusion will lead to the loss of key features, which will seriously affect the recognition performance and become an important challenge in this field. To solve this problem, the current research mainly focuses on two aspects: one is to improve the face recognition algorithm itself to make it more robust to occlusion; The second is to restore the complete face through occlusion repair technology, and then perform recognition. This paper systematically summarizes the latest progress of occluded face recognition technology. First, it analyzes the direct recognition methods, then discusses the methods based on occlusion repair, and compares the advantages and disadvantages of the two strategies and their applicable scenarios. Finally, this paper looks forward to the future research directions and provides a reference for the further development of occluded face recognition technology.
文章引用：王晓彤, 韩金颖, 王俊兴, 李振乾. 遮挡人脸识别技术综述[J]. 计算机科学与应用, 2025, 15(5): 651-659. https://doi.org/10.12677/csa.2025.155137

参考文献

[1] He, L., Li, H., Zhang, Q. and Sun, Z. (2019) Dynamic Feature Matching for Partial Face Recognition. IEEE Transactions on Image Processing, 28, 791-802. [Google Scholar] [CrossRef] [PubMed]
[2] Huber, M., Boutros, F., Kirchbuchner, F. and Damer, N. (2021) Mask-Invariant Face Recognition through Template-Level Knowledge Distillation. 2021 16th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2021), Jodhpur, 15-18 December 2021, 1-8. [Google Scholar] [CrossRef] 
[3] Alashbi, A., Mohamed, A.H.H.M., El-Saleh, A.A., Shayea, I., Sunar, M.S., Alqahtani, Z.R., et al. (2025) Human Face Localization and Detection in Highly Occluded Unconstrained Environments. Engineering Science and Technology, an International Journal, 61, Article 101893. [Google Scholar] [CrossRef] 
[4] Song, L., Gong, D., Li, Z., Liu, C. and Liu, W. (2019) Occlusion Robust Face Recognition Based on Mask Learning with Pairwise Differential Siamese Network. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, 27 October 2019-2 November 2019, 773-782. [Google Scholar] [CrossRef] 
[5] 路杨, 康瑞梅, 张芳君. 分块稀疏表示的人脸识别研究[J]. 软件工程与应用, 2016, 5(5): 277-284.
[6] 王海. 基于分块稀疏表示的部分遮挡人脸识别[J]. 信息与电脑, 2019, 31(22): 64-65.
[7] 狄岚, 矫慧文, 梁久祯. 稀疏综合字典学习的小样本人脸识别[J]. 智能系统学报, 2021, 16(2): 218-227.
[8] Gao, G., Yang, J., Jing, X., Shen, F., Yang, W. and Yue, D. (2017) Learning Robust and Discriminative Low-Rank Representations for Face Recognition with Occlusion. Pattern Recognition, 66, 129-143. [Google Scholar] [CrossRef] 
[9] Zhang, W. and Chen, X. (2025) Robust Discriminative Non-Negative and Symmetric Low-Rank Projection Learning for Feature Extraction. Symmetry, 17, Article 307. [Google Scholar] [CrossRef] 
[10] Gao, G., Huang, P., Zhou, Q., Hu, Z. and Yue, D. (2017) Low-Rank Representation and Locality-Constrained Regression for Robust Low-Resolution Face Recognition. In: Lu, H. and Xu, X., Eds., Studies in Computational Intelligence, Springer International Publishing, 17-26. [Google Scholar] [CrossRef] 
[11] Li, H., Zhang, Y., Wang, W., Zhang, S. and Zhang, S. (2024) Recovery-Based Occluded Face Recognition by Identity-Guided Inpainting. Sensors, 24, Article 394. [Google Scholar] [CrossRef] [PubMed]
[12] Man, Q. and Cho, Y. (2023) Efficient Face Region Occlusion Repair Based on T-GANs. Electronics, 12, Article 2162. [Google Scholar] [CrossRef] 
[13] Xu, X., Sarafianos, N. and Kakadiaris, I.A. (2020) On Improving the Generalization of Face Recognition in the Presence of Occlusions. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, 14-19 June 2020, 3470-3480. [Google Scholar] [CrossRef] 
[14] Yu, C., Wang, J., Peng, C., Gao, C., Yu, G. and Sang, N. (2018) BiseNet: Bilateral Segmentation Network for Real-Time Semantic Segmentation. In: Ferrari, V., Hebert, M., Sminchisescu, C. and Weiss, Y., Eds., Lecture Notes in Computer Science, Springer International Publishing, 334-349. [Google Scholar] [CrossRef] 
[15] 罗健, 郝泳涛. 基于GAN的人脸图像三维重建[J]. 电脑知识与技术, 2022, 18(23): 71-73+83.