Reversible Data Hiding under Inconsistent Distortion Metrics.

Area of Science:

• Digital watermarking and steganography
• Information theory and coding

Background:

• Recursive code construction (RCC) based on optimal transition probability matrix (OTPM) approaches the rate-distortion bound for reversible data hiding (RDH).
• Existing methods for OTPM estimation are limited to consistent distortion metrics, where host elements share the same distortion measure.
• Many real-world applications involve position-dependent, inconsistent distortion metrics, often quantified as multi-distortion metrics.

Purpose of the Study:

• To formulate the rate-distortion problem for RDH under multi-distortion metrics.
• To propose a general framework for estimating OTPM with inconsistent distortion metrics.
• To extend RCC to approach the rate-distortion bound for RDH under multi-distortion metrics.

Main Methods:

• Formulation of the rate-distortion problem for RDH with multi-distortion metrics.
• Development of a general framework for estimating OTPM under inconsistent distortion metrics.
• Application of the framework to RDH in color images and reversible steganography.

Main Results:

• The proposed framework successfully estimates OTPM for inconsistent distortion metrics.
• RCC extended with the new framework effectively approaches the rate-distortion bound for RDH under multi-distortion metrics.
• Demonstrated improvements over existing techniques in RDH for color images and reversible steganography.

Conclusions:

• The developed framework provides a generalized approach for RDH under multi-distortion metrics.
• This method enhances the performance of RCC, enabling it to better approach theoretical limits in complex scenarios.
• The findings are applicable to various RDH applications with position-dependent distortions.