CTU-Row Geometric Transforms for Local Texture Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face limitations in optimizing the orientation of coding tree units (CTUs) within a picture, leading to suboptimal compression performance due to fixed or uniform orientation assumptions that do not account for varying local texture patterns within an input picture.
Innovation Solution
Applying geometric transforms, such as horizontal and vertical flips, and rotations, to groups of CTUs or GTUs within a picture, allowing different groups to have distinct orientations based on their local texture patterns, thereby enhancing compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed or uniform orientation assumptions are applied to all CTUs in a picture, then device complexity is reduced and processing is simplified, but compression performance deteriorates due to inability to account for varying local texture patterns
Solution Approach 1:
The picture is divided into multiple groups of CTUs, where each group can have its own orientation transform applied independently. This segmentation allows different regions with different texture patterns to be processed with appropriate orientations, improving compression performance while keeping the complexity manageable through localized processing rather than global complexity.
Solution Approach 2:
Different orientation transforms are applied to different groups of CTUs based on their local texture patterns. Instead of applying a uniform orientation assumption across the entire picture, the patent applies local quality adaptation where each group's orientation is optimized for its specific content characteristics, thereby improving compression performance without requiring excessive complexity.
2Productivity
If different geometric transforms are applied to different groups of CTUs based on local texture patterns, then compression performance is improved, but device complexity increases due to multiple transform operations
Solution Approach 1:
The orientation transform is made dynamic by selecting different geometric transforms for different groups of CTUs based on their texture patterns. This dynamic adaptation allows the system to optimize compression performance for each local region while managing overall complexity through selective application of transforms only where beneficial, rather than applying complex transforms uniformly across the entire picture.
Solution Approach 2:
The patent changes the orientation parameter (geometric transform type) for different groups of CTUs based on local texture characteristics. By varying the transform parameters locally rather than maintaining a fixed parameter across the entire picture, the system achieves improved compression performance while controlling complexity through parameter adaptation rather than structural complexity.
3Loss of substance
If geometric transforms are applied to adapt to local texture patterns, then compression efficiency is improved and data volume is reduced, but processing time increases due to additional transform operations
Solution Approach 1:
The picture is segmented into groups of CTUs that can be processed independently with appropriate geometric transforms. This segmentation enables parallel processing of different groups, reducing overall processing time while still achieving the compression benefits of localized orientation adaptation. The segmented approach allows data volume reduction through efficient local transforms without requiring sequential processing of the entire picture.
Solution Approach 2:
Geometric transforms are applied selectively to groups of CTUs where they provide the most benefit, rather than applying them uniformly to the entire picture. This partial action approach reduces processing time by avoiding unnecessary transforms in regions where they would not improve compression, while still achieving significant data volume reduction in the regions where transforms are applied.
Data Source
AI summary
A video bitstream comprising a current picture of a video is received. A first group of samples and a second group of samples in the current picture are determined. A first geometric transform is determined for the first group of samples in the current picture and a second geometric transform is determined for the second group of samples in the current picture. The first geometric transform is configured to adjust an orientation of the first group of samples in the current picture. The second geometric transform is different from the first geometric transform and configured to adjust an orientation of the second group of samples in the current picture. The picture is reconstructed, where the first group of samples is reconstructed based on the determined first geometric transform and the second group of samples is reconstructed based on the determined second geometric transform.