| 引用本文： |
-
王志轩,王洁,谢聪格,李睿.基于数据拆分的图像隐私保护方案[J].信息安全学报,已采用 [点击复制]
- wang zhixuan,Wang Jie,Xie congge,Li Rui.Image privacy protection scheme based on data splitting[J].Journal of Cyber Security,Accept [点击复制]
|
|
| 摘要: |
| 随着云存储技术的广泛应用，图像隐私保护已成为云环境中的研究热点。然而，现有工作多侧重于隐私保护机制本身，在一定程度上忽视了图像在实际应用场景下的可用性需求。因此，如何在保障隐私安全的前提下实现图像的有效可用，仍是该领域面临的核心挑战。针对上述问题，本文提出了一种基于数据拆分的图像隐私保护方案，在有效保护图像隐私信息的同时支持原始图像的安全重构，满足了隐私保护和图像的使用需求。该方案通过将图像拆分为主图碎片（公共区域）与子图碎片（隐私区域），并为不同类型的图像碎片建立相应索引，分别进行分布式存储，从而实现隐私信息的有效隔离。在此基础上，为支持图像在授权场景下的安全重构，进一步提出一种索引关联算法，用于在多云环境中高效建立隐私图像碎片与公共图像碎片之间的映射关系。此外，为适应碎片化图像数据的动态变化特性，设计了一种索引动态管理机制，实现对索引的增加与删除。同时，为保障图像碎片在存储与传输过程中的完整性，防止数据被篡改，引入水印嵌入与验证机制。在重构阶段对图像碎片来源与完整性进行校验，保证图像重构过程的安全可信。在真实图像数据集上的实验结果表明，所提方案能够在有效保护图像隐私的前提下，实现高效的图像拆分存储与可靠重构，在图像安全性与可用性上展现出显著优势。 |
| 关键词: 隐私保护 图像拆分 索引关联 验证重构 |
| DOI： |
| 投稿时间：2026-01-14修订日期：2026-07-18 |
| 基金项目:国家自然科学基金项目（面上项目，重点项目，重大项目）,国家重点基础研究发展计划（973计划） |
|
| Image privacy protection scheme based on data splitting |
|
wang zhixuan, Wang Jie, Xie congge, Li Rui
|
| (东莞理工学院) |
| Abstract: |
| With the widespread application of cloud storage technology, image privacy protection has become a research hotspot in the cloud environment. However, existing work focuses on the privacy protection mechanism itself, neglecting the usability requirements of images in practical application scenarios to some extent. Therefore, how to achieve effective image usability while ensuring privacy and security remains a core challenge in this field. To address this issue, this paper proposes an image privacy protection scheme based on data decomposition, which effectively protects image privacy information while supporting secure reconstruction of the original image, meeting both privacy protection and image usage requirements. This scheme decomposes the image into main image fragments (public areas) and sub-image fragments (privacy areas), and establishes corresponding indexes for different types of image fragments, storing them separately in a distributed manner to achieve effective isolation of privacy information. Based on this, to support secure reconstruction of images in authorized scenarios, an index association algorithm is further proposed to efficiently establish the mapping relationship between privacy image fragments and public image fragments in a multi-cloud environment. In addition, to adapt to the dynamic changes of fragmented image data, a dynamic index management mechanism is designed to enable the addition and deletion of indexes. Simultaneously, to ensure the integrity of image fragments during storage and transmission and prevent data tampering, a watermark embedding and verification mechanism is introduced. The source and integrity of image fragments are verified during the reconstruction phase to ensure the security and reliability of the image reconstruction process. Experimental results on real image datasets demonstrate that the proposed scheme can achieve efficient image splitting and storage and reliable reconstruction while effectively protecting image privacy, exhibiting significant advantages in image security and usability. |
| Key words: privacy protection image splitting index association validation reconstruction |
