Video Block Vector Prediction for Content-Adaptive Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies, such as MPEG-2, MPEG-4, AVC, HEVC, and VVC, require improvements in coding efficiency and effectiveness for video processing.
Innovation Solution
Implementing block vector prediction (BVP) and intra block copy (IBC) merge modes based on content type, where block vector differences (BVD) are adjusted for natural and screen contents, enhancing coding efficiency by inheriting motion information and applying appropriate offsets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video coding technologies (MPEG-2, MPEG-4, AVC, HEVC, VVC) are used, then video processing can be performed with standard methods, but coding efficiency and effectiveness are insufficient
Solution Approach 1:
The patent applies different BVD offset values tailored to specific content types (natural content vs. screen content). By identifying the content type and selecting appropriate offsets locally, the system optimizes coding efficiency for each content category without compromising overall coding effectiveness across diverse video materials.
Solution Approach 2:
The patent changes the BVD offset parameter based on content type classification. Different offset values are applied depending on whether the content is natural or screen content, allowing the system to adapt parameters dynamically to improve both coding efficiency and effectiveness for different video categories.
2Productivity
If a single BVD offset is used for all content types, then the processing method is simple, but coding efficiency is not optimized for different content characteristics
Solution Approach 1:
The system implements content-type-specific BVD offsets where different offset values are applied to natural content and screen content separately. This local differentiation optimizes coding efficiency for each content type while maintaining a relatively simple overall structure through automated content classification.
Solution Approach 2:
The BVD offset parameter is dynamically changed based on detected content type. The system automatically selects appropriate offset values without requiring complex manual configuration, balancing improved coding efficiency with acceptable processing complexity through automated parameter adaptation.
3Reliability
If motion information is not inherited from BVP candidates, then the conversion process is straightforward, but coding effectiveness is reduced
Solution Approach 1:
The system performs preliminary determination of motion information from BVP candidates before the actual video block conversion. By pre-processing and inheriting motion information in advance, the system improves coding effectiveness while keeping the main conversion process relatively simple through the use of pre-computed motion data.
Solution Approach 2:
The BVP candidates automatically provide motion information that is inherited by the current video block during conversion. This self-service mechanism allows the system to improve coding effectiveness by utilizing motion information from neighboring blocks without requiring complex external motion estimation processes.
Data Source
AI summary
Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. In the method, for a conversion between a current video block of a video and a bitstream of the video, motion information of a block vector prediction (BVP) candidate of the current video block is determined. The conversion is performed based on the motion information, wherein a block vector (BV) is inherited from the motion information.