On the Security of Bitstream-level JPEG Encryption with Restart Markers

TL;DR

The paper evaluates the security of a bitstream-level JPEG encryption method that uses restart markers ($RI$) to enable partial encryption and block permutation while preserving JPEG compatibility and file size. It expands security assessment beyond key space to ciphertext-only attacks, employing key-space bounds, key-sensitivity, NZCA, and histogram analyses. Results show that RST markers can improve robustness, with security dependent on the restart interval and careful parameter choices. The findings support the method's advantage over conventional approaches for COA resistance, while highlighting trade-offs linked to RI settings and extended-block handling.

Abstract

This paper aims to evaluate the security of a bitstream-level JPEG encryption method using restart (RST) markers, where encrypted image can keep the JPEG file format with the same file size as non-encrypted image. Data encrypted using this method can be decoded without altering header information by employing a standard JPEG decoder. Moreover, the use of RST markers enables the definition of extended blocks divided by the markers, so spatially partial encryption and block-permutation-based encryption can be carried out. However, the security of the method was evaluated only with respect to the key space analysis for brute-force attacks and other limited attacks. Accordingly, in this paper, we evaluated the security of the method with respect to robustness against ciphertext-only attacks including state-of-the-art attacks. In experiments, the method is compared with conventional encryption methods, and it is confirmed to be robust against ciphertext-only attacks if parameters used for image encryption are carefully chosen.

Figures (20)

• Figure 1: Type 3
• Figure 2: Type 4
• Figure 4: Structure of JPEG bitstream.
• Figure 5: Structure of image data.
• Figure 6: Block diagram of encryption process.