1 College of Information Engineering, North China University of Water Resources and Electric Power, Zhengzhou 450046, China
2 College of Electronic Information, Sichuan University, Chengdu 610065, China
Show less
文章历史+
Received
Published
Revised
2025-08-02
2026-07-08
2026-01-09
Online Date
2026-08-06
PDF (7639K)
摘要
云环境中加密图像检索能够有效保护数据隐私,但在密文状态下特征表达受限、视觉语义模糊仍是限制检索精度的关键瓶颈。提出一种融合动态特征表达与跨模态语义协同约束的加密图像检索方法。首先,构建非对称交叉注意力驱动的局部-全局动态特征融合网络,通过关联建模实现细粒度纹理与全局语义的互补表达,显著增强视觉特征的判别能力。其次,设计视觉-文本语义约束的跨模态哈希协同机制,将融合特征与文本特征映射至统一哈希空间,通过文本对视觉特征的弱监督纠偏,提高跨模态一致性与哈希码可分性。最后,结合BFV(Brakerski-Fan-Vercauteren)同态加密模型,实现密文哈希码的相似度计算,使检索在加密状态下完成,确保特征与结果的隐私安全。在Olivia-2688、GHIM-10k、Caltech-101及NWPU-RESISC45四个公开数据集上进行实验验证。结果表明,所提出的动态融合模块使mAP(mean average precision) 提升14.01%~15.23%,跨模态语义约束进一步提升4.56%~9.93%;同时,多项式环阶数为8 192时,密文相似度计算与明文结果保持完全一致,且具备较高的计算效率。综合结果表明,该方法能够在复杂场景下实现高判别特征表达,实现了云环境中跨模态特征协同的高精度加密图像检索。
Abstract
In cloud environments, encrypted image retrieval can effectively protect data privacy; however, limited feature expressiveness and weakened visual semantics in the ciphertext domain remain major challenges that constrain retrieval accuracy. To address these issues, an encrypted image retrieval method that integrates dynamic feature representation with cross-modal semantic co-constraints was proposed. First, a non-symmetric cross-attention-based local-global dynamic feature fusion network was developed to model the complementary relationships between fine-grained textures and global semantics, thereby enhancing the discriminability of visual features. Second, a visual-text semantic constrained cross-modal hashing framework was designed to project fused visual features and textual features into a unified hashing space, where textual semantics provide weak supervisory guidance to reduce visual ambiguity and improve cross-modal consistency. Third, the BFV(Brakerski-Fan-Vercauteren) homomorphic encryption scheme was incorporated to enable similarity computation directly in the encrypted domain, ensuring the privacy of both feature representations and retrieval results. Experiments conducted on Olivia-2688, GHIM-10k, Caltech-101, and NWPU-RESISC45 demonstrate that the proposed dynamic fusion module improves mAP(mean average precision) by 14.01%~15.23%, while the cross-modal semantic constraint yields an additional 4.56%~9.93% performance gain. Furthermore, when the polynomial ring degree is set to 8 192, ciphertext-domain similarity computation is fully consistent with its plaintext counterpart and exhibits high computational efficiency. The results confirm that the proposed method achieves highly discriminative encrypted feature representation and supports high-precision encrypted image retrieval in complex cloud environments.
智能终端设备的普及使得数据量激增,用户越来越倾向于将数据存储在云服务器中,这也带来了隐私泄露的风险。基于内容的图像检索(content based image retrieval, CBIR)技术是云存储环境中广泛应用的重要支撑,传统的明文存储与检索模式已无法满足隐私保护需求,尤其是在医疗诊断、安防等高数据安全需求的应用场景中。为此,研究者通过加密算法确保图像在传输、存储以及检索过程中的安全性,同时利用先进的特征提取技术从图像中提取出有效特征信息,构建索引并进行相似度计算,从而实现高效、准确的隐私保护图像检索[1-3]。
FarayolaO A, OlorunfemiO L, ShoetanP O. Data privacy and security in it: a review of techniques and challenges[J]. Computer Science & IT Research Journal, 2024, 5(3): 606-615.
[2]
ZhangB, XuY Y, YanY J, et al. Privacy-preserving image retrieval based on additive secret sharing in cloud environment[J]. Cluster Computing, 2024, 27(4): 5021-5045.
[3]
FengQ H, LiP Y, LuZ X, et al. Evit: privacy-preserving image retrieval via encrypted vision transformer in cloud computing[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2024, 34(8): 7467-7483.
[4]
梁海华. 面向隐私保护的加密图像检索研究[D]. 上海: 上海大学, 2020.
[5]
LiangHaihua. Research on encrypted image retrieval for privacy protection[D]. Shanghai: Shanghai University, 2020.
[6]
MartoliaM, DhanoreN, SinghA, et al. A modified local binary pattern (LBP) for content-based image retrieval[J]. International Journal of Advanced Science and Technology, 2020, 29(1): 1630-1644.
[7]
XiaZ H, WangL, TangJ, et al. A privacy-preserving image retrieval scheme using secure local binary pattern in cloud computing[J]. IEEE Transactions on Network Science and Engineering, 2020, 8(1): 318-330.
[8]
MaW T, ZhouT Q, QinJ H, et al. A privacy-preserving content-based image retrieval method based on deep learning in cloud computing[J]. Expert Systems with Applications, 2022, 203: 117508.
[9]
ZhangX Y, HongJ W. A dynamic authorizable ciphertext image retrieval algorithm based on security neural network inference[J]. PloS One, 2024, 19(10): e0309947.
[10]
WangY, ChenL Q, WuG, et al. Efficient and secure content-based image retrieval with deep neural networks in the mobile cloud computing[J]. Computers & Security, 2023, 128: 103163.
[11]
SongC, HuangR W. Secure convolution neural network inference based on homomorphic encryption[J]. Applied Sciences, 2023, 13(10): 6117.
[12]
ZhangQ, YanY, LinY, et al. Image security retrieval based on chaotic algorithm and deep learning[J]. IEEE Access, 2022, 10: 67210-67218.
[13]
WangZ, ZhangQ Y, MengL T, et al. Secure content based image retrieval scheme based on deep hashing and searchable encryption[J]. Computers, Materials & Continua, 2023, 75(3): 6161-6184.
[14]
YangT F, MaJ F, MiaoY B, et al. Muteir: traceable encrypted image retrieval in the multi-user setting[J]. IEEE Transactions on Services Computing, 2022, 16(2): 1282-1295.
[15]
LiY Y, MaJ F, MiaoY B, et al. Similarity search for encrypted images in secure cloud computing[J]. IEEE Transactions on Cloud Computing, 2020, 10(2): 1142-1155.
[16]
WuY R, GuoH F, ChakrabortyC, et al. Edge computing driven low-light image dynamic enhancement for object detection[J]. IEEE Transactions on Network Science and Engineering, 2022, 10(5): 3086-3098.
PanHuafeng, WangChunling, WuTao. Research on image retrieval recommendation based on hash network and sensitive hashing[J]. Computer Technology and Development, 2022, 32(7): 173-178.
[19]
ChenH Z, ZhouH, ZhangJ, et al. Perceptual hashing of deep convolutional neural networks for model copy detection[J]. ACM Transactions on Multimedia Computing, Communications and Applications, 2023, 19(3): 1-20.
[20]
LuoX, WangH X, WuD Q, et al. A survey on deep hashing methods[J]. ACM Transactions on Knowledge Discovery from Data, 2023, 17(1): 1-50.
[21]
XiaoX N, CaoS, WangL J, et al. Deep hashing image retrieval based on hybrid neural network and optimized metric learning[J]. Knowledge-Based Systems, 2024, 284: 111336.
[22]
JiangX, TangH, LiZ H. Global meets local: dual activation hashing network for large-scale fine-grained image retrieval[J]. IEEE Transactions on Knowledge and Data Engineering, 2024, 36(11): 6266-6279.
[23]
MohammedZ K, MohammedM A, AbdulkareemK H, et al. A metaverse framework for IoT-based remote patient monitoring and virtual consultations using AES-256 encryption[J]. Applied Soft Computing, 2024, 158: 111588.
[24]
BosJ, DucasL, KiltzE, et al. CRYSTALS-Kyber: a CCA-secure module-lattice-based KEM[C]// 2018 IEEE European Symposium on Security and Privacy (EuroS&P). Piscataway: IEEE, 2018: 353-367.
LiTao, ZhuWenzhong, CheXuan. Image recognition of apple leaf disease based on improved ResNet-50 and transfer learning[J]. Science Technology and Engineering, 2024, 24(24): 10370-10381.
[27]
DaiK Y, ShaoJ, GongB, et al. CLIP-FSSC: a transferable visual model for fish and shrimp species classification based on natural language supervision[J]. Aquacultural Engineering, 2024, 107: 102460.
MaRongheng, YuChunyu, TongYixin. Image denoising based on residual dense networks and attention mechanisms[J]. Science Technology and Engineering, 2025, 25(9): 3795-3805.
[30]
MaZ Y, WangS W, LuoX, et al. HARR: learning discriminative and high-quality hash codes for image retrieval[J]. ACM Transactions on Multimedia Computing, Communications and Applications, 2024, 20(5): 1-23.
[31]
SrivastavaP, KhareA. Utilizing multiscale local binary pattern for content-based image retrieval[J]. Multimedia Tools and Applications, 2018, 77: 12377-12403.
[32]
ChuK, LiuG H. Image retrieval based on a multi-integration features model[J]. Mathematical Problems in Engineering, 2020, 2020(1): 1461459.
[33]
LiF F, FergusR, PeronaP. Learning generative visual models from few training examples: an incremental bayesian approach tested on 101 object categories[C]// 2004 Conference on Computer Vision and Pattern Recognition Workshop. Piscataway: IEEE, 2004: 178.
[34]
ChengG, HanJ W, LuX Q. Remote sensing image scene classification: benchmark and state of the art[J]. Proceedings of the IEEE, 2017, 105(10): 1865-1883.
[35]
PanW Y, WangM M, QinJ H, et al. Improved CNN-based hashing for encrypted image retrieval[J]. Security and Communication Networks, 2021(1): 5556634.
[36]
DubeyS R, SinghS K, ChuW T. Vision transformer hashing for image retrieval[C]// Proceedings of the 2022 IEEE International Conference on Multimedia and Expo (ICME). Piscataway: IEEE, 2022: 1-6.
[37]
GuoZ, WeiB X, CaiQ L, et al. POST: prototype-oriented similarity transfer framework for cross-domain facial expression recognition[J]. Computer Animation and Virtual Worlds, 2024, 35(3): e2260.
[38]
TianM, SuM T, XiaoX L, et al. High-precision privacy-protected image retrieval based on multi-feature fusion[J]. Knowledge-Based Systems, 2025, 315: 113243.