Reversible Data Hiding via Histogram Modification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data hiding techniques are not reversible, leading to permanent distortion of the cover media and limited payload capacity, which is inadequate for applications requiring lossless data extraction and high visual quality.
Innovation Solution
A reversible data embedding technique that modifies the histogram of an image by shifting pixel values around zero and peak points to embed data, allowing for a large payload while maintaining high visual quality, with a peak signal-to-noise ratio (PSNR) of at least 48 dB.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If spread-spectrum based data hiding methods are used, then data embedding capability is improved, but reversibility deteriorates due to truncation error and round-off error
Solution Approach 1:
The patent changes the parameter of pixel value modification from arbitrary modifications (spread-spectrum) to controlled histogram-based modifications. By modifying only histogram bins corresponding to zero or minimum points and peak points, the method achieves reversibility while maintaining payload capacity through systematic parameter adjustment rather than random modification.
Solution Approach 2:
The patent performs preliminary histogram analysis to identify zero/minimum points and peak points before data embedding. This preliminary action enables the system to prepare the image structure in advance, creating designated regions for data embedding that can be systematically reversed later without affecting other image regions.
2Ease of manufacture
If least significant bit-plane approach is used, then data embedding is simplified, but losslessness deteriorates due to bit-replacement without memory
Solution Approach 1:
The patent introduces histogram modification as an intermediary mechanism between data embedding and image modification. Instead of directly modifying pixel values (as in LSB approach), the method uses histogram bin adjustment as an intermediate step, which provides a structured way to embed data while maintaining the ability to reverse the process through systematic histogram inversion.
3Quantity of substance
If quantization-index-modulation is used, then data embedding capability is improved, but distortion-free quality deteriorates due to quantization error
Solution Approach 1:
The patent applies local quality modification by targeting specific local regions in the histogram (zero/minimum points and peak points) rather than applying uniform modification across the entire image. This localized approach minimizes overall distortion while maximizing data embedding capacity in specific regions where histogram modification has minimal visual impact.
4Reliability
If modulo 256 addition is used for authentication, then reversibility is improved, but visual quality deteriorates due to salt-and-pepper noise
Solution Approach 1:
The patent applies partial action by modifying only specific histogram bins (those corresponding to zero/minimum points and peak points) rather than applying modification to all pixel values. This partial modification approach maintains reversibility through systematic changes while minimizing visual artifacts by leaving the majority of histogram bins unchanged.
Data Source
AI summary
Methods and apparatus are provided for encoding a pixel domain image with hidden data by modifying the histogram of the pixel domain image to make space for such hidden data.