High-Capacity Reversible Data Hiding in Encrypted Images Based on MSB Bit-Plane Serpentine Rearrangement and Compression

Reversible data hiding techniques in encrypted images (RDH-EI) have been extensively researched to safeguard the privacy of images. To enhance the embedding capability of RDH-EI algorithms and security, this work leverages the topological coherence within the input plaintext image to create more additional space for embedding confidential information prior to image encryption. Initially, the prediction error of the input image is computed using the pixel prediction technique. Subsequently, the bit-plane serpentine rearrangement mechanism and the bit-stream compression method are employed to further enhance the process. This compression technique efficiently compresses the most significant bit-planes, resulting in a high-capacity space for embedding. Finally, the end-user successfully extracts the data and restores the input image using the possessed keys. The results clearly show that the proposed approach achieves not only attains a superior embedding capacity which achieves the maximum average embedding rate with a value of 3.624 (bpp), but also obtains a commendable reconstructed image quality that is PSNR=+∞ and SSIM=1.