Video Processing Intra Mode Generation for Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in improving coding efficiency, particularly in handling video units with intra template matching (TM) and intra copy block (IBC) modes, as well as screen content coding.
Innovation Solution
The proposed method involves generating an intra mode for video units based on coding information associated with intra TM or IBC coded blocks, and determining which coding tools to disable for screen content coding, thereby optimizing the conversion between video units and bitstreams.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video coding techniques are used for intra template matching and intra copy blocks, then the coding process is simpler, but coding efficiency is lower
Solution Approach 1:
The patent dynamically determines intra mode based on video unit characteristics and coding information. The encoder adaptively selects between intra template matching mode and intra copy mode depending on the content type (e.g., screen content vs. natural video), allowing the coding process to adjust its complexity to match the actual requirements of each video unit, thereby improving overall coding efficiency without unnecessarily increasing complexity for all cases.
Solution Approach 2:
The patent changes key coding parameters including disabling certain coding tools for screen content coding, adjusting transform skip mode, and modifying residual coding parameters based on the detected video unit type. These parameter changes optimize the coding process for specific content types, improving compression performance while maintaining manageable complexity through targeted adjustments rather than universal complexity increases.
2Productivity
If all coding tools are enabled for screen content coding, then more coding options are available, but coding efficiency is reduced due to unnecessary processing
Solution Approach 1:
The patent extracts and disables specific coding tools that are unnecessary for screen content coding. By identifying and removing irrelevant coding operations from the processing pipeline, the patent reduces unnecessary computational overhead and improves coding efficiency for screen content while maintaining the availability of essential coding tools when needed.
Solution Approach 2:
The patent applies different coding tool configurations to different regions or types of video content. Screen content regions receive a optimized subset of coding tools, while other regions may utilize the full range of tools. This localized approach to tool availability ensures that each region receives the appropriate level of processing complexity, improving overall efficiency without sacrificing versatility where required.
Data Source
AI summary
Embodiments of the disclosure provide a solution for video processing. A method for video processing is proposed. The method includes: generating, for a conversion between a video unit of a video and a bitstream of the video, an intra mode for the video unit based on coding information associated with the video unit, wherein the video unit is an intra template matching (TM) coded block or the video unit is an intra copy block (IBC) coded block; and performing the conversion based on the generated intra mode.