Geometry Transformation Adaptive Loop Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current adaptive loop filter (ALF) and geometry transformation-based ALF (GALF) designs in video codecs face limitations such as restricted filter support selection, limited geometric transformations, and inadequate chroma component filtering, leading to suboptimal coding performance, especially in cases with unclear directional information or smooth regions.
Innovation Solution
The implementation of multiple filter supports and expanded geometric transformations, along with block-level control for chroma components, allows for more flexible filtering and the use of luma component information in chroma filtering, enhancing the decoding and encoding processes in advanced video codecs like HEVC.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If only one pre-defined filter support is selected for each video unit, then device complexity is reduced, but coding performance is limited
Solution Approach 1:
The patent divides the video unit into multiple regions, each with its own filter support selection. Instead of applying a single filter support to the entire video unit, the encoder selects different filter supports for different regions based on local characteristics, thereby improving coding performance without excessively increasing overall complexity.
Solution Approach 2:
The patent enables different filter supports to be applied to different regions within a video unit based on local image characteristics. This allows the filtering operation to adapt to local variations in the image content, improving coding performance by matching the filter support to the specific requirements of each region.
2Adaptability or versatility
If current geometric transformations with only 90 degree changes are used, then device complexity is reduced, but adaptability to different directional information is limited
Solution Approach 1:
The patent introduces dynamic geometric transformations that can adapt to the directional characteristics of the image content. Instead of being limited to fixed 90-degree transformations, the system dynamically selects and applies transformations (including 45-degree rotations and flips) based on the detected gradient directions and image characteristics, thereby improving adaptability to different directional information.
Solution Approach 2:
The patent changes the parameters of geometric transformations by introducing additional transformation types beyond the standard 90-degree changes. Specifically, it incorporates 45-degree rotations and flip operations, allowing the filtering process to better align with various directional patterns in the image content.
3Measurement precision
If classification is based only on gradients, then directional information is captured well, but smooth regions and regions with progressive row changes cannot be differentiated
Solution Approach 1:
The patent combines multiple classification criteria (gradients, smoothness measures, and row-based variations) to create a composite classification system. This composite approach allows the system to differentiate between various types of regions including directional regions, smooth regions, and regions with progressive row changes, thereby improving both measurement precision and adaptability.
Solution Approach 2:
The patent creates a multi-functional classification system that can handle multiple types of image regions using a unified approach. The classification mechanism is designed to detect and classify different region types (directional, smooth, progressive row changes) within a single framework, improving versatility without requiring separate specialized systems for each region type.
4Manufacturing precision
If a single 5x5 diamond-shaped filter support is used for chroma components, then device complexity is reduced, but coding performance and chroma filtering quality are limited
Solution Approach 1:
The patent applies different filter supports to different chroma components (Cb and Cr) and different regions based on their specific characteristics. Instead of using a uniform 5x5 diamond-shaped filter for all chroma components, the system selects appropriate filter supports for each component and region, thereby improving chroma filtering quality by matching the filter characteristics to the specific requirements of each chroma region.
Solution Approach 2:
The patent segments the chroma filtering process by applying different filter supports to different chroma components and regions. This segmentation allows the system to optimize filtering for each specific chroma region rather than applying a single generic filter, thereby improving overall chroma filtering quality.
Data Source
AI summary
The present disclosure provides various techniques related to adaptive loop filtering (ALF), and particular to geometry transformation-based ALF (GALF). In an aspect, a method for decoding video data includes receiving an encoded bitstream having coded video data from which reconstructed video units are generated, identifying multiple filter supports for the reconstructed video units, and filtering the reconstructed video units using the respective multiple filter supports to produce a decoded video output. Another method includes enabling block-level control of ALF of chroma components for the reconstructed video units, performing, for the reconstructed video units, the block-level ALF for the chroma components when ALF is enabled for one video block and skip performing the block-level ALF for the chroma components when ALF is disabled for another video block, and generating, based on the enabled block-level control of ALF, a decoded video output. Related devices, means, and computer-readable medium are also described.