Subblock Template Matching for Accurate Motion Vector Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in predicting motion vectors at a subblock level, leading to increased data transmission overhead and reduced coding efficiency.
Innovation Solution
The implementation of subblock-based template-matching for motion vector prediction (SbTMVP) within a current template, where motion vectors for subblocks are derived based on a center position MV and adjacent subblock MVs, enhancing prediction accuracy and reducing transmission overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vectors are predicted at block level using existing technologies, then the coding process is simpler, but the prediction accuracy is insufficient leading to increased data transmission overhead
Solution Approach 1:
The current block is divided into multiple subblocks, and motion vectors are derived separately for each subblock using template matching. This segmentation allows for more precise motion representation in different regions of the block, improving prediction accuracy while maintaining manageable complexity through systematic processing of smaller units
Solution Approach 2:
Different motion vectors are computed for different subblocks within the current block based on their respective template matching results. This local quality approach ensures that each subblock has motion information optimized for its specific region, capturing local motion variations that block-level prediction would miss
2Measurement precision
If subblock-based motion vector prediction is implemented, then the prediction accuracy improves, but the data transmission overhead increases
Solution Approach 1:
Motion vectors for all subblocks are derived in advance during the decoding process using template matching with previously decoded subblocks as references. This preliminary action allows the decoder to reconstruct motion information without receiving it explicitly in the bitstream, reducing transmission overhead while maintaining high prediction accuracy
Solution Approach 2:
The system uses already-decoded subblocks within the current picture as reference templates for deriving motion vectors of other subblocks. This self-service mechanism eliminates the need to transmit reference motion information separately, as the decoder generates it autonomously from available data, reducing data transmission requirements
Data Source
AI summary
A video bitstream is received. The video bitstream includes a current block comprising a plurality of subblocks and a template region of the current block comprising a plurality of template subblocks adjacent to at least one of a top side and a left side of the current block. A motion vector (MV) located in a center position of the current block is determined. The MV is determined based on at least one MV of the plurality of subblocks of the current block. A MV for each of the plurality of template subblocks is determined based on the MV located in the center position of the current block and a respective MV of a corresponding subblock of the plurality of subblocks that is adjacent to the respective template subblock. The current block is reconstructed based on the determined MVs for the plurality of template subblocks.