Adaptive Color Transform for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current color transform schemes in video coding introduce signaling redundancy, overlapping coding tools, and high buffer size requirements, particularly in scenarios involving the H.265/HEVC and emerging VVC standards.
Innovation Solution
The method involves obtaining a coded video bitstream in RGB format, decoding signaling information for a current coding unit, determining adaptive color transform enablement, and converting residual information from the YCgCo domain to the RGB domain when ACT is enabled, to reconstruct the coding unit based on prediction information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If color transform is performed in video coding, then coding efficiency is improved, but signaling redundancy increases
Solution Approach 1:
The patent applies parameter changes by dynamically switching between different color transform modes (ACT enabled/disabled) based on content characteristics. The encoder determines whether to apply adaptive color transform to coding units based on chroma prediction mode and other parameters, changing the transform parameter adaptively rather than using a fixed mode for all blocks, thereby improving coding efficiency while controlling signaling overhead
Solution Approach 2:
The patent segments the video picture into coding units and applies color transform selectively at the CU level rather than globally. By dividing the picture into smaller coding units and applying ACT only to specific CUs where it benefits coding efficiency, the patent reduces overall signaling redundancy while maintaining improved compression where applicable
2Productivity
If adaptive color transform is applied to all coding units, then coding efficiency is improved, but device complexity increases
Solution Approach 1:
The patent implements partial action by applying adaptive color transform only to specific coding units where it provides benefit, rather than to all CUs. The encoder uses conditions such as chroma prediction mode and ACT enablement flags to determine selective application, performing the transform partially on only those blocks where it improves efficiency, thereby reducing overall processing complexity
Solution Approach 2:
The patent introduces dynamics by making the color transform application adaptive and conditional rather than static. The ACT enablement information and chroma prediction modes allow the system to dynamically adjust whether transform is applied based on local content characteristics, making the processing flexible and adaptable to different scene requirements
3Measurement precision
If color transform is performed, then chroma component representation is improved, but buffer size requirements increase
Solution Approach 1:
The patent applies local quality by using different color transform representations locally in different coding units based on their specific characteristics. Rather than maintaining high-precision transformed chroma data globally, the system applies ACT locally where needed and uses conventional representation elsewhere, reducing overall buffer requirements while maintaining improved chroma representation where beneficial
Data Source
AI summary
A method for controlling color component processing for a decoder includes: obtaining a coded video bitstream, the video bitstream being coded from a source having a RGB color format; decoding, signaling information of a current coding unit (CU) in a segment of a current picture from the coded video bitstream, the signaling information includes prediction information of the current CU and a control signal for processing chroma components of the current CU in a prediction residual domain; determining, according to the control signal, residual information of the current CU, comprising: determining, according to the control signal, adaptive color transform (ACT) enablement information; and when the ACT enablement information indicates that ACT is used for the current CU, obtaining the residual information by converting residuals at YCgCo domain obtained from inverse transform to residuals at RGB domain; and reconstructing the current CU based on the residual information and the prediction information.