Transform Unit Sign Bit Hiding with Multi-Set Coefficient Parity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding standards, such as H.264/AVC and developing MPEG-H, spend a significant portion of the bitstream on encoding sign bits for non-zero coefficients, which can be optimized to reduce data overhead.
Innovation Solution
Implement multi-level significance maps and parity-based sign bit hiding techniques for subsets of coefficients within transform units, particularly for larger units like 16x16 and 32x32, to hide sign bits by adjusting coefficients to match parity values, reducing the need for explicit sign bit encoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sign bits are explicitly encoded for each non-zero coefficient, then the sign information is accurately transmitted, but the bitstream size increases significantly
Solution Approach 1:
The patent extracts the sign bit information from the explicit encoding stream and embeds it within the coefficient data itself. By taking out the separate sign bit signaling and integrating it into the coefficient magnitude encoding, the patent eliminates redundant signaling while preserving sign information accuracy.
Solution Approach 2:
The patent merges the sign bit transmission with the coefficient magnitude encoding process. Instead of separate signaling, the sign information is combined with the magnitude data through the parity-based embedding mechanism, reducing overall bitstream volume while maintaining information完整性.
2Quantity of substance
If sign bit hiding is applied to all coefficients, then bitstream size is minimized, but decoding complexity increases
Solution Approach 1:
The patent segments the transform unit into multiple subsets and applies sign bit hiding selectively based on local characteristics. By dividing the coefficient block into regions with different hiding strategies, the patent reduces overall decoding complexity while maintaining compression efficiency.
Solution Approach 2:
The patent applies sign bit hiding partially rather than universally - specifically to coefficients meeting certain criteria (non-zero values in specific positions). This partial application reduces the computational burden on the decoder compared to universal sign bit hiding.
3Quantity of substance
If coefficients are adjusted to match parity values, then sign bits are successfully hidden, but manufacturing precision of coefficient values changes
Solution Approach 1:
The patent introduces parity checking as an intermediary mechanism to verify that coefficient adjustments have not corrupted the embedded sign information. This mediator ensures that even though coefficients are modified, the original sign bits can be reliably recovered during decoding.
Solution Approach 2:
The patent changes the parameter encoding scheme from explicit sign bits to parity-based implicit encoding. By transforming the representation method rather than the underlying value, the patent hides sign information without significantly altering the perceptual quality of the reconstructed coefficients.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods of encoding and decoding for video data are described for encoding or decoding coefficients for a transform unit. In particular, the sign bits for the non-zero coefficients are encoded using sign bit hiding. Two or more sets of coefficients are defined for the transform unit and a sign bit may be hidden for each set, subject to satisfaction of a threshold test. The sets may correspond to coefficient groups that are otherwise used in multi-level significance map encoding and decoding.