Securing color medical images using data hiding method based on modified Octa-PVD and hyperchaotic system

Abstract

This paper proposes an enhanced data-hiding method using PVD that combines LSB and a hyperchaotic system to embed textual data inside a color medical image without causing any observable image distortions. We first select the host color image and then compress it by leveraging the rapid and an effective vector quantization (VQ) method to ensure relevance and fidelity during the data-hiding process. Based on the joint permutation and diffusion (JPD) system, a hyperchaotic system encrypts the compressed image blocks to safeguard the concealed image data from unauthorized access during the data-hiding process. The LSB replacement conceals the text data within the transformed central pixel blocks. Afterward, determine directly the following outer pixels for embedding using Octa-PVD for each direction where the values of branch conditions (BCs) are smaller. Otherwise, it uses LSB insertion. The secret message is extracted directly from the stego image without referencing the host image, which preserves the stego image's quality and reduces extraction time. The experimental results demonstrate that the proposed method outperforms traditional PVD and multidimensional PVD-based methods. It has significantly higher PSNR and SSIM values and RS and PDH exhibiting resilience against.