Visually Meaningful Image Encryption Scheme Using Multi-Parameter Fractal Theory and Block Synchronous Sorting Diffusion
Zhihua Gan,Mengge Sun,Yalin Song,Xiuli Chai,Donghua Jiang,Guoqiang Long,Xin He
DOI: https://doi.org/10.1088/1402-4896/ace28b
2023-01-01
Physica Scripta
Abstract:Traditional visually meaningful image encryption (VMIE) works by embedding a secret image (SI) into a visually meaningful carrier image (CAI), thus achieving the dual protection of both data information and appearance features. However, the current VMIE algorithm still suffers from problems of reconstruction quality and transmission efficiency. To address these issues, this paper proposes an innovative VMIE algorithm that utilizes a newly designed two-dimensional hyperchaotic map, multi-parameter fractal matrix (MPFM) theory and compressive sensing. The proposed algorithm achieves dual protection of both the semantic information and appearance image data. First, a newly designed 2D infinite triangle folding map (2D-ITFM) is presented to generate a binary key-controlled measurement matrix to measure and compress a plain image (PI) to generate measured image (MI) and decrease the amount of transmission information. Next, based on the hyperchaotic map 2D-ITFM and MPFM, we present the inter-block scrambling (IBS) algorithm and intra-block synchronous sorting diffusion (IBSSD) algorithm, which are used to process the MI and generate SI. Then, the matrix encoding embedding (MEE) technique is utilized to hide the SI into the CAI to produce a visually secure cipher image (CII). Finally, the experimental results demonstrate that our scheme is effective in improving the anti-attack ability while guaranteeing good imperceptibility and reconstruction performance. This scheme can be employed in the field of information security communication.