[Paper Review] Reversible Data Hiding in Encrypted Images Based on Bit-plane Compression of Prediction Error

Youqing Wu, Wenjing Ma+3 more|arXiv (Cornell University)|Jul 8, 2020

Advanced Steganography and Watermarking Techniques32 references4 citations

TL;DR

This paper proposes a high-capacity reversible data hiding in encrypted images (RDHEI) method that leverages joint bit-plane compression of prediction error to maximize embedding capacity. By rearranging and jointly encoding bit-planes of prediction errors before encryption, the method reserves more space for data embedding, achieving up to 3.763 bpp average embedding rate on standard databases, outperforming state-of-the-art methods.

ABSTRACT

As a technology that can prevent the information from being disclosed, the reversible data hiding in encrypted images (RDHEI) acts as an important role in privacy protection and information security. To make use of the image redundancy and further improve the embedding performance, a high-capacity RDHEI method based on bit-plane compression of prediction error is proposed in this paper. Firstly, the whole prediction error is calculated and divided into blocks of the same size. Then, the content owner rearranges the bit-plane of prediction error by block and compresses the bitstream with the joint encoding algorithm to reserve room. Finally, the image is encrypted and the information can be embedded into the reserved room. On the receiver side, the information extraction and the image recovery are performed separably. Experimental results show that the proposed method brings higher embedding capacity than state-of-the-art RDHEI works.

Motivation & Objective

• To address the limited embedding capacity in existing reversible data hiding in encrypted images (RDHEI) methods.
• To preserve image privacy by ensuring the original image remains confidential even when data is embedded in the ciphertext.
• To enable separable information extraction and image recovery using different keys, enhancing practical applicability.
• To exploit image redundancy more effectively through bit-plane compression of prediction error.
• To improve embedding capacity by reserving more room in the encrypted domain through efficient joint encoding.

Proposed method

• The method computes the prediction error for each pixel in the original image using a predictor.
• The prediction error is divided into blocks, and the bit-planes of each block are rearranged to enhance compression efficiency.
• A joint encoding algorithm—combining Huffman and run-length encoding—is applied to compress the bit-planes, minimizing bitstream size.
• The compressed bitstream is used to reserve space in the image before encryption, creating room for data embedding.
• The image is then encrypted using a symmetric encryption scheme, preserving the reserved space for embedding.
• On the receiver side, the embedded data is extracted first, followed by image recovery using the same key and decompression process.

Experimental results

Research questions

• RQ1 Can joint bit-plane compression of prediction error significantly increase the embedding capacity in RDHEI?
• RQ2 How does the proposed method compare to state-of-the-art RDHEI techniques in terms of embedding rate and reversibility?
• RQ3 To what extent does the joint encoding algorithm improve compression efficiency compared to separate encoding?
• RQ4 Can the method achieve separable data extraction and image recovery without compromising security or capacity?
• RQ5 Does the method maintain high image quality and perfect reversibility after embedding and decryption?

Key findings

• The proposed method achieves an average embedding rate of 3.763 bpp on the BOSSBase database, surpassing the best existing methods by 0.081 bpp.
• On the BOWS-2 database, the average embedding rate reaches 3.642 bpp, an improvement of 0.147 bpp over the state of the art.
• For the Lena and Baboon test images, the method achieves embedding rates of 3.15 bpp and 1.526 bpp, respectively, demonstrating high capacity.
• The method ensures perfect reversibility, with MSE = 0 and SSIM = 1, confirming exact image recovery.
• The joint encoding algorithm effectively compresses bit-planes by leveraging both short and long bit-string patterns, outperforming separate encoding strategies.
• The method maintains high performance even in complex textures, with minimal impact from invalid pixels (e.g., reference and overflow pixels).