Reversible data hiding in encrypted images using LWE-based secret sharing

Area of Science:

• Cryptography and Information Security
• Digital Image Processing
• Data Hiding Techniques

Background:

• Reversible data hiding in encrypted images (RDH-EI) is crucial for secure data embedding without compromising image confidentiality.
• Existing RDH-EI methods often struggle with accommodating multiple data hiders effectively.
• The Learning With Errors (LWE) problem offers robust security foundations for cryptographic applications.

Purpose of the Study:

• To introduce a novel RDH-EI technique capable of supporting multiple data hiders.
• To enhance embedding capacity and ensure lossless image reconstruction even with data loss.
• To leverage the security of the LWE problem within an RDH-EI framework.

Main Methods:

• Integration of secret sharing based on the Learning With Errors (LWE) problem with adaptive coding strategies.
• Distribution of the original image to multiple data hiders using LWE-based methods.
• Application of block permutation and stream cipher encryption to preserve spatial correlation.
• Adaptive compression of most significant bits (MSB) blocks based on frequency for increased capacity.
• Embedding of extra data combined with inverse Huffman codewords and auxiliary information into MSB planes.

Main Results:

• The proposed RDH-EI method successfully accommodates multiple data hiders.
• Lossless restoration of the original image is achievable even with damaged or missing data shares, given sufficient valid shares.
• Experimental results show superior performance compared to state-of-the-art methods, particularly in embedding capacity.
• The technique demonstrates robust security due to its foundation on the LWE problem.

Conclusions:

• The novel RDH-EI technique effectively addresses the challenge of multiple data hiders.
• The integration of LWE-based secret sharing and adaptive coding significantly boosts embedding capacity.
• The method provides a secure and reliable solution for reversible data hiding in encrypted images.