Adaptive Loop Filter Signaling for 360-Degree Video Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In-loop filtering across discontinuous boundaries in 360-degree video processing results in poor visual quality and decreased coding efficiency due to the joint processing of referenced pixels across these boundaries, leading to visible face seam artifacts.
Innovation Solution
The proposed method signals Adaptive Loop Filter (ALF) processing using Adaptive Parameter Set (APS) indices, allowing for the selection of ALF filter sets and applying them to blocks within Coding Tree Blocks (CTBs), and when processing across virtual boundaries, replacing outside samples with padded samples to avoid using uncorrelated data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If in-loop filtering is applied across discontinuous boundaries in 360-degree video, then filtering coverage is improved, but visual quality deteriorates due to face seam artifacts
Solution Approach 1:
The patent segments the video picture into multiple tiles with discontinuous boundaries, and applies different filtering strategies to different regions. Specifically, it prevents in-loop filtering across virtual boundaries that separate different faces in 360-degree video, while allowing filtering within continuous regions. This segmentation approach maintains filtering coverage in valid areas while avoiding artifact generation at face boundaries.
Solution Approach 2:
The patent applies local quality control by differentiating between regions where filtering should be applied and regions where it should be prevented. It uses virtual boundary detection to identify discontinuous edges between faces, and selectively disables filtering operations across these specific boundaries while maintaining filtering in continuous regions. This local differentiation ensures high visual quality at face boundaries while preserving filtering benefits in other areas.
2Reliability
If in-loop filtering is applied across discontinuous boundaries, then filtering effectiveness is improved, but coding efficiency decreases
Solution Approach 1:
The patent segments the filtering operation into regions across virtual boundaries and regions within continuous areas. By preventing filtering across virtual boundaries and applying it within continuous regions, the patent achieves effective filtering where it matters while avoiding redundant or harmful operations at boundaries, thus improving coding efficiency without sacrificing filtering effectiveness in critical areas.
Solution Approach 2:
The patent applies partial filtering action by selectively enabling filtering only in regions where it produces beneficial effects. Instead of applying filtering uniformly across the entire picture including across discontinuous boundaries, it applies filtering partially - specifically within continuous regions and excluding boundary regions. This partial action approach maintains filtering effectiveness in valid areas while reducing overall processing overhead and improving coding efficiency.
3Measurement precision
If ALF processing uses outside samples across virtual boundaries, then filtering accuracy is improved, but artifact generation increases
Solution Approach 1:
The patent extracts and removes outside samples that would cause artifacts when used across virtual boundaries. It identifies samples located across discontinuous boundaries and excludes them from filtering operations, using only samples within continuous regions. This extraction of problematic samples prevents artifact generation while maintaining filtering accuracy using only correlated samples from within the same face.
Solution Approach 2:
The patent applies preliminary anti-action by preventing the use of outside samples across virtual boundaries before artifacts can be generated. It proactively identifies virtual boundaries and blocks access to samples across these boundaries, thereby preventing the harmful effect of artifact generation rather than correcting it afterward. This preliminary prevention maintains filtering accuracy using only safe, correlated samples.
Data Source
AI summary
According to a method for Adaptive Loop Filter (ALF) processing of reconstructed video, multiple indicators are signaled in slice at an encoder side or parsed at a decoder side, where the multiple indicators are Adaptive Parameter Set (APS) indices associated with temporal ALF filter sets for the ALF processing. A current indicator is determined from the multiple indicators, where the current indicator is used to select a current ALF filter set. Filtered-reconstructed pixels are derived for the current block by applying the current ALF filter to the current block. In another method, if the ALF processing applied at a target sample requires an outside sample on other side of a target virtual boundary from the target sample, the outside sample is replaced by a padded sample.


