SECURE REVERSABLE DATA HIDING USING EDGE AWARE AND MULTI MSB SELF-PREDICTION
DOI:
https://doi.org/10.62643/Abstract
The increasing adoption of cloud computing has elevated the significance of reversible data hiding in encrypted images (RDH-EI) as a key technology for safeguarding data security and privacy. This paper presents an advanced RDH-EI technique that leverages a dual prediction strategy, combining edge-directed prediction and multi-bit self-prediction of the most significant bits (MSBs). Initially, the image owner enhances pixel prediction accuracy by utilizing an edge-directed approach based on least-square optimization. Subsequently, self prediction is applied to each pixel, starting from the MSB and progressing adaptively through higher-order bit-planes, one bit-plane at a time. A binary location map is generated to track the prediction errors for each bit-plane. These location maps, after being loss lessly compressed using the Joint Bi-level Image Experts Group (JBIG) algorithm, are embedded into the encrypted image as auxiliary data, thus providing additional space for embedding extra data. The data hider retrieves this auxiliary data and embeds new information into the reserved space, creating a marked encrypted image. Authorized recipients, equipped with specific keys, can either perfectly recover the embedded data or loss lessly restore the original image from the marked encrypted image without any degradation. Experimental results demonstrate that the proposed method achieves impressive pure embedding rates of 4.240 bpp for the BOSS Base dataset and 4.088 bpp for the BOWS-2 dataset, outperforming many current RDH-EI techniques in terms of embedding capacity and accuracy.
Downloads
Published
Issue
Section
License

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.













