Video Motion Vector Prediction With Fixed 4×4 Subblocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards face inefficiencies in motion vector prediction, particularly in the construction of candidate lists and sub-CU size settings, leading to redundant operations and mismatched storage granularities in Advanced Temporal Motion Vector Prediction (ATMVP) and Temporal Motion Vector Prediction (TMVP) technologies.
Innovation Solution
A method and device for video image processing that divides coding tree units into coding units using a quadtree plus binary tree division, scales motion vectors based on reference images, and optimizes the construction of motion vector candidate lists to reduce redundant operations and improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ATMVP technology divides current CU into N×N sub-blocks with adaptive size setting, then prediction accuracy is improved, but device complexity increases due to multiple size configurations and redundant operations
Solution Approach 1:
The current CU is divided into multiple sub-blocks with a fixed 4×4 configuration rather than adaptive N×N sizing. This segmentation approach maintains prediction accuracy by processing smaller units while simplifying the complexity of size configuration and reducing redundant operations through a standardized division method.
Solution Approach 2:
The patent changes the parameter of sub-block size from adaptive (4, 8, 16, 32, 64, 128) to fixed (4×4). This parameter simplification reduces device complexity by eliminating the need for multiple size configurations while maintaining prediction effectiveness through the standardized 4×4 sub-block structure.
2Measurement precision
If motion vector candidate list is constructed with multiple temporal and spatial candidates, then prediction accuracy is improved, but processing time increases due to scanning and selecting optimal candidates
Solution Approach 1:
The patent uses partial action by limiting the motion vector candidate list to a fixed number of candidates (up to 4) with specific temporal and spatial configurations. This partial construction approach provides sufficient prediction accuracy without the excessive processing time required to scan and select from larger candidate sets, achieving an optimal balance between accuracy and efficiency.
3Adaptability or versatility
If sub-CU size is adaptively set at frame level, then flexibility is improved, but storage granularity mismatch occurs with motion information storage format
Solution Approach 1:
The patent changes the sub-CU size parameter from adaptive (4, 8, 16, 32, 64, 128) to fixed (4×4), which aligns with the storage granularity of motion information. This parameter change ensures that the sub-CU size matches the motion information storage format, eliminating granularity mismatches while maintaining flexibility through the standardized configuration.
Data Source
AI summary
A video image decoding method includes receiving a bitstream that includes an index indicating a combination of two motion vector offsets, determining a base motion vector list that includes a dual-prediction base motion vector group including a first base motion vector and a second base motion vector, and determining the two motion vector offsets from a preset offset set according to the index included in the bitstream. The two motion vector offsets correspond to the first base motion vector and the second base motion vector, respectively. The method further includes determining motion vectors of a current image block according to the first base motion vector, the second base motion vector, and the two motion vector offsets, and performing a prediction on the current image block according to the motion vectors of the current image block.