<p>In cloud storage systems, ensuring both privacy and usability of encrypted images remains a challenging task. Conventional encryption schemes either eliminate visual information completely or expose partial content, failing to offer a satisfactory balance. Thumbnail-Preserving Encryption has recently emerged as a promising solution by preserving the coarse visual structure of images while hiding fine details. This paper presents a practical framework compatible with both JPEG and PNG formats. For JPEG images, we propose a novel Tunable Sum-Preserving Interval-Constrained Encryption (TSICE) scheme, which regulates the range of encrypted DC coefficients through a tunable parameter <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(T\)</EquationSource> <EquationSource Format="MATHML"><math> <mi>T</mi> </math></EquationSource> </InlineEquation>, while preserving the global sum of DC coefficients. For PNG images, we make a small improvement to an existing sum-preserving scheme to enhance the avalanche effect. In both cases, the ciphertext and plaintext share the same global sum, which is leveraged as plaintext-related input for key derivation. This property works on both substitution and permutation stages without requiring external side information. We further investigate thumbnail generation procedures adopted by real-world platforms such as Google Drive and evaluate the visual usability of encrypted thumbnails through simulation. Extensive experiments demonstrate that the proposed framework achieves a practical trade-off between privacy protection and visual usability.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A practical chaos-based framework for sum-preserving JPEG encryption in cloud storage

  • Ziwen Zhu,
  • Geng Zhao,
  • Yingjie Ma

摘要

In cloud storage systems, ensuring both privacy and usability of encrypted images remains a challenging task. Conventional encryption schemes either eliminate visual information completely or expose partial content, failing to offer a satisfactory balance. Thumbnail-Preserving Encryption has recently emerged as a promising solution by preserving the coarse visual structure of images while hiding fine details. This paper presents a practical framework compatible with both JPEG and PNG formats. For JPEG images, we propose a novel Tunable Sum-Preserving Interval-Constrained Encryption (TSICE) scheme, which regulates the range of encrypted DC coefficients through a tunable parameter \(T\) T , while preserving the global sum of DC coefficients. For PNG images, we make a small improvement to an existing sum-preserving scheme to enhance the avalanche effect. In both cases, the ciphertext and plaintext share the same global sum, which is leveraged as plaintext-related input for key derivation. This property works on both substitution and permutation stages without requiring external side information. We further investigate thumbnail generation procedures adopted by real-world platforms such as Google Drive and evaluate the visual usability of encrypted thumbnails through simulation. Extensive experiments demonstrate that the proposed framework achieves a practical trade-off between privacy protection and visual usability.