Sign Data Hiding in Image Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding methods face challenges in efficiently adjusting coefficient values within coefficient groups to optimize compression, particularly in cases where the sum of coefficients is odd or even and the sign of the first non-zero coefficient is plus or minus, leading to incorrect sign representation during sign data hiding.
Innovation Solution
A method and system for encoding data that selects a line to define an adjustment target coefficient group within each coefficient group, calculates the sum of coefficients, and adjusts the value of one adjustment target coefficient based on the sum and sign of the first non-zero coefficient, without scanning for the last non-zero coefficient, to ensure correct sign representation and efficient compression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If scanning for the last non-zero coefficient is performed to adjust coefficient values, then sign representation accuracy is improved, but hardware overhead and processing complexity increase
Solution Approach 1:
The patent extracts and processes only the essential information (sum of coefficients and its parity) needed for sign data hiding, eliminating the need for complex scanning operations to find the last non-zero coefficient. By taking out only the necessary computational elements, the solution achieves sign representation accuracy without the hardware overhead of full scanning.
Solution Approach 2:
The sum of coefficients is calculated preliminarily before the sign data hiding process, and its parity is determined in advance. This preliminary calculation provides all necessary information for subsequent coefficient adjustment, eliminating the need for additional scanning operations and reducing hardware complexity while maintaining accuracy.
2Measurement precision
If scanning for the last non-zero coefficient is performed, then correct sign adjustment is achieved, but processing speed decreases
Solution Approach 1:
The patent extracts only the essential computational elements (coefficient sum and its parity) needed for sign adjustment, eliminating time-consuming scanning operations. This extraction approach maintains sign adjustment correctness while significantly improving processing speed by removing unnecessary computational steps.
Solution Approach 2:
The coefficient sum and its parity are calculated in advance before the sign data hiding process begins. This preliminary action provides all necessary information for correct sign adjustment, eliminating the need for subsequent scanning operations and thereby improving processing speed without sacrificing accuracy.
3Productivity
If coefficient values are adjusted based on sum parity, then sign data hiding efficiency is improved, but coefficient value precision may be affected
Solution Approach 1:
The patent applies local quality by making minimal, targeted adjustments to coefficient values (changing only by ±1) rather than large-scale modifications. This localized adjustment approach maintains the overall precision of coefficient values while achieving the necessary sign data hiding efficiency through small, controlled changes.
Solution Approach 2:
The patent changes the parameter of coefficient values by minimal amounts (±1) based on the parity of the sum, rather than making large changes. This subtle parameter adjustment achieves sign data hiding efficiency while preserving the precision of the original coefficient values as much as possible.
Data Source
AI summary
A method of encoding data includes selecting a line to define an adjustment target coefficient group in each of a plurality of coefficient groups included in a transform unit that has been transformed and quantized. Each of the coefficient groups comprises a plurality of coefficients. For each of the coefficient groups, a sum of the coefficients for the respective coefficient group is calculated. For each of the coefficient groups, a value of one adjustment target coefficient included in the adjustment target coefficient group is adjusted according to a result of the calculation of the sum of the coefficients for the respective coefficient group.