Rate and Distortion Optimization for Reversible Data Hiding Using Multiple Histogram Shifting.

Area of Science:

• Digital Image Processing
• Information Security
• Computer Science

Background:

• Reversible data hiding using histogram shifting (HS) is crucial for stego-image quality.
• Conventional HS methods struggle with optimal selection of side information (peak and zero bins) due to computational complexity and large solution spaces.
• Empirical criteria in existing schemes often fail to achieve globally optimal reversible embedding.

Purpose of the Study:

• To formulate HS-based multiple embedding as a rate and distortion optimization problem.
• To develop a novel scheme for adaptively determining optimal peak and zero bins for enhanced reversible data hiding.
• To improve embedding capacity and stego-image quality in HS-based reversible data hiding.

Main Methods:

• Developed a rate and distortion model for HS-based multiple embedding.
• Derived two key propositions for efficient distortion computation and solution space reduction.
• Employed a genetic algorithm to search for nearly optimal zero and peak bins.

Main Results:

• The proposed scheme adaptively determines the number and values of peak and zero bin pairs for a given data payload.
• Experimental results show superior performance in embedding capacity compared to previous approaches.
• The method achieves better stego-image quality than existing HS-based reversible embedding schemes.

Conclusions:

• The developed optimization framework provides a superior approach to HS-based multiple reversible data hiding.
• The genetic algorithm effectively searches for optimal parameters, overcoming limitations of empirical methods.
• This research significantly advances the field of reversible data hiding by improving both capacity and quality.