Video Coding Bit Allocation for Padded Slice Quality Balance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media encoding and decoding methods result in uneven subjective quality of padded pictures due to differences in reconstruction quality between slices with and without padding content, leading to inconsistent visual experience.
Innovation Solution
Adjust coding parameters based on full padding information to balance the number of coded bits across coding units, ensuring consistent reconstruction quality by adjusting the coding units with excessive padding to match the bit allocation of other units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If picture padding is performed to ensure integer number of slices, then the encoder and decoder can operate normally, but the subjective quality of the padded picture becomes uneven
Solution Approach 1:
The patent applies local quality by differentiating the treatment of fully padded coding units from non-fully padded coding units. Fully padded coding units (where all samples are padding samples) are identified and allocated fewer coded bits, while non-fully padded coding units receive more coded bits. This localized differentiation in bit allocation resolves the quality uniformity issue by ensuring that padding regions do not consume excessive coding resources that would otherwise be available for actual picture content.
Solution Approach 2:
The patent changes the coding parameter (number of coded bits) based on the padding status of coding units. By introducing a flag or condition that identifies whether a coding unit is fully padded, the system dynamically adjusts the bit allocation parameter. This parameter change allows the encoding system to allocate bits more efficiently, preventing the quality degradation caused by uniform bit allocation across all coding units including padding regions.
2Ease of manufacture
If a padded picture is encoded with uniform bit allocation, then encoding is simple, but reconstruction quality differs between slices with padding and slices without padding
Solution Approach 1:
The patent implements local quality by applying different encoding strategies to different regions of the picture. Specifically, coding units that are fully padded are identified and assigned a different number of coded bits compared to non-fully padded coding units. This localized approach maintains encoding simplicity through automated detection while improving reconstruction quality consistency by ensuring that padding regions do not unfairly compete for coding resources with actual picture content.
Solution Approach 2:
The patent segments the picture into coding units and further identifies which coding units are fully padded versus non-fully padded. This segmentation allows the encoding system to apply different bit allocation rules to different segments, thereby resolving the quality inconsistency issue while maintaining a relatively simple encoding process through systematic classification and differential treatment of segments.
3Productivity
If fully padded coding units occupy fewer coded bits, then bitstream compression is improved, but non-fully padded coding units have more abundant bits leading to quality imbalance
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the number of coded bits allocated to each coding unit based on its padding status. Fully padded coding units are assigned fewer bits, while non-fully padded coding units are assigned more bits. This parameter adjustment resolves the quality balance issue by ensuring that the bit allocation reflects the actual information content of each coding unit, preventing quality imbalance while maintaining compression efficiency.
Solution Approach 2:
The patent implements local quality by applying different bit allocation parameters to different coding units based on their local characteristics (whether they are fully padded or not). This localized quality approach ensures that coding units with actual picture content receive sufficient bits for high-quality reconstruction, while fully padded coding units occupy fewer bits, thereby achieving both compression efficiency and quality balance.
Data Source
Figure 1a
Figure 1b
Figure 2
AI summary
Embodiments of this application relate to the field of media technologies, and disclose an encoding method, a decoding method, and an apparatus, to balance subjective quality of a padded picture. The method includes: performing picture padding on a to-be-encoded picture to obtain a padded picture; then, obtaining a coding unit based on the padded picture; next, determining coding information of the coding unit based on full padding information of the coding unit; and afterwards, encoding the coding unit based on the coding information to generate a bitstream. The full padding information indicates whether all samples in the coding unit are picture padding samples, and the coding information includes at least one of a number of padding bits or a coding length.