LMCS Chroma Scaling Without Redundant Clipping in Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The current design of Luma Mapping with Chroma Scaling (LMCS) in video coding standards like VVC incurs extra computation complexity and on-chip memory requirements due to redundant clipping during domain conversions, affecting coding efficiency.
Innovation Solution
The proposed method involves converting reconstructed luma samples from a mapped domain to an original domain without clipping, and using these converted samples to derive chroma scaling factors for decoding, thereby eliminating redundant clipping and reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant clipping is performed during domain conversions in LMCS, then video quality is maintained, but computational complexity and on-chip memory requirements increase
Solution Approach 1:
The patent extracts and removes the redundant clipping operation from the domain conversion process. By identifying that clipping is only necessary in specific cases (when mapping between different bit depths with different ranges), the patent eliminates the universal clipping operation that was previously applied in all domain conversions, thereby reducing computational complexity while maintaining video quality where needed
Solution Approach 2:
The patent applies clipping selectively rather than uniformly across all samples. The clipping operation is applied only to specific samples that fall outside the valid range after domain conversion, rather than applying clipping to all samples. This localized approach maintains video quality for affected samples while avoiding unnecessary computation for samples that are already within valid ranges
2Reliability
If redundant clipping is performed during domain conversions in LMCS, then video quality is maintained, but on-chip memory requirements increase
Solution Approach 1:
The patent removes the redundant clipping operation from the processing pipeline, which eliminates the need for additional buffer memory that would be required to store intermediate clipped values. By performing clipping only when necessary and integrating it into the existing processing flow, the patent reduces on-chip memory requirements while maintaining video quality
3Manufacturing precision
If domain conversions with clipping are performed, then coding accuracy is maintained, but processing speed decreases
Solution Approach 1:
The patent applies clipping only partially - specifically to samples that require it - rather than applying clipping excessively to all samples. By using conditional logic to determine whether clipping is needed for each sample based on its value and the target bit depth requirements, the patent maintains coding accuracy for samples that need clipping while avoiding the processing overhead for samples that don't require clipping, thereby improving overall processing speed
Data Source
AI summary
Methods and devices are provided for reducing the decoding latency introduced by luma mapping with chroma scaling (LMCS). In one method, during decoding of a coding unit (CU) that is coded by an inter mode or Combined Inter and Intra Prediction (CIIP) mode under LMCS framework, a plurality of reconstructed samples of luma component is obtained in a mapped domain, a plurality of converted samples of luma component is obtained in an original domain by converting the plurality of reconstructed samples of luma component from the mapped domain into the original domain, and the plurality of converted samples of luma component is used in the original domain, without clipping, in deriving chroma scaling factors for decoding the chroma samples of the CU.