Cross-Component Adaptive Loop Filtering Padding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face challenges in efficiently handling cross-component adaptive loop filtering, particularly at virtual boundaries, where padding methods can be sub-optimal and inefficient, and different padding techniques are required for various boundaries and 360-degree video coding.
Innovation Solution
The implementation of cross-component adaptive loop filtering (CC-ALF) with modified padding methods, such as mirrored and repetitive padding, is introduced to improve filtering efficiency across various boundaries, including ALF virtual boundaries, picture/subpicture/slice/tile boundaries, and 360-degree video virtual boundaries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional padding methods are used for cross-component adaptive loop filtering at virtual boundaries, then the filtering process can be implemented, but the filtering efficiency is sub-optimal and inefficient
Solution Approach 1:
The patent applies different padding methods (mirrored padding, repetitive padding) to different boundary types (picture boundaries, slice boundaries, tile boundaries, virtual boundaries) based on their specific characteristics. This local differentiation optimizes filtering efficiency for each boundary type while maintaining overall processing effectiveness.
Solution Approach 2:
The patent dynamically selects padding methods based on the boundary type and filtering context. The system adapts the padding approach (mirrored vs. repetitive) according to the specific virtual boundary characteristics, enabling efficient processing without uniform application of a single method.
2Manufacturing precision
If different padding techniques are applied for various boundaries and 360-degree video coding, then the quality of reconstructed video is improved, but the device complexity increases
Solution Approach 1:
The patent segments the video processing into different boundary types (picture, slice, tile, virtual boundaries) and applies appropriate padding methods to each segment. This segmentation allows complex processing to be divided into manageable, boundary-specific operations that improve reconstruction quality without overwhelming system complexity.
Solution Approach 2:
The patent changes the padding parameter (method type) based on the boundary characteristics and video type (including 360-degree video). By adjusting the padding approach according to specific parameters, the system achieves high reconstruction quality while managing complexity through parameter-based adaptation rather than structural complexity.
3Ease of operation
If mirrored padding process is applied for padding unavailable luma sample, then the filtering operation can proceed, but inefficiencies occur at virtual boundaries
Solution Approach 1:
The patent dynamically switches between mirrored padding and repetitive padding based on the boundary type. At virtual boundaries, the system adapts to use the more efficient repetitive padding method, while maintaining mirrored padding where appropriate, thus optimizing both operational continuity and filtering efficiency.
Data Source
AI summary
A method for video processing is described. The method includes determining, for a conversion between a video unit of a video and a bitstream representation of the video, whether to enable a mirrored padding process for padding an unavailable luma sample during an application of a loop filtering tool to the video unit; and performing the conversion based on the determining.