Intra Block Copy Motion Candidate Limiting for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in improving coding efficiency, particularly in reducing the complexity of subsequent processes while maintaining effective video compression.
Innovation Solution
A method for video processing that determines a target number for a set of motion candidates, limited by an upper limit value, and uses a fast rate distortion optimization (RDO) process to select the set of motion candidates for an intra block copy (IBC) based mode, thereby reducing the complexity of subsequent processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the number of motion candidates is not limited, then more comprehensive motion prediction can be achieved, but the complexity of subsequent processing increases
Solution Approach 1:
The patent applies partial action by selecting only the top N motion candidates (where N is a predetermined number smaller than the total number of available candidates) based on rate distortion cost. This partial selection approach maintains sufficient prediction accuracy while significantly reducing processing complexity, as the system only needs to evaluate and process the top N candidates rather than all possible motion candidates.
Solution Approach 2:
The patent changes the parameter of motion candidate selection by introducing an upper limit N on the number of candidates to be processed. This parameter modification transforms the unbounded search space into a bounded one, enabling the system to achieve a balance between prediction reliability and processing efficiency by controlling the search depth and computational load.
2Productivity
If fast rate distortion optimization is applied without upper limit, then better coding efficiency can be achieved, but processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-determining the upper limit N for motion candidate selection before the rate distortion optimization process. This preliminary constraint setting allows the fast RDO process to operate within a bounded search space from the outset, achieving good coding efficiency while preventing excessive processing time accumulation that would occur with unlimited candidate evaluation.
Solution Approach 2:
The patent implements the skipping principle by rapidly selecting only the top N motion candidates with the lowest rate distortion costs and skipping the detailed processing of remaining candidates. This rushing through approach maintains sufficient coding efficiency by focusing computational resources on the most promising candidates while dramatically reducing overall processing time compared to exhaustive evaluation.
Data Source
AI summary
Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: determining, for a conversion between the current video block and a bitstream of the video, a target number for a set of motion candidates, the target number being smaller than or equal to an upper limit value; determining, based on the target number and a fast rate distortion optimization (RDO) process, the set of motion candidates from a plurality of motion candidates for an intra block copy (IBC) based mode for coding the current video block; and performing the conversion based on the set of motion candidates.


