Sub-Block Temporal Motion Vector Prediction for Efficient Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently encoding and decoding high-definition and ultra-high-definition video data while maintaining image quality, particularly in terms of reducing redundancy and optimizing motion vector prediction.
Innovation Solution
Implementing sub-block motion vector prediction by determining a motion shift vector for a current coding unit based on spatial neighbor blocks and reconstructing sub-block-based temporal motion vectors from co-located pictures, utilizing computing devices and non-transitory computer readable storage media to execute programs for efficient video encoding and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If traditional block-based motion vector prediction is used, then the encoding process is simpler, but the amount of data required for motion information increases and coding efficiency decreases
Solution Approach 1:
The current block is divided into multiple sub-blocks, and each sub-block has its own motion vector prediction. This segmentation allows for more precise motion representation in different regions of the block, reducing the overall data required while maintaining coding efficiency. The patent applies this by creating sub-block temporal motion vector predictions that are specific to each sub-region rather than using a single prediction for the entire block.
Solution Approach 2:
Different sub-blocks within the same block may have different motion characteristics, and the patent applies local quality by allowing each sub-block to have its own temporal motion vector prediction derived from corresponding sub-blocks in co-located blocks. This localizes the prediction quality to match local motion variations, reducing the need for additional data to represent these variations.
2Productivity
If sub-block temporal motion vector prediction is implemented, then coding efficiency improves and data required for motion information reduces, but the encoding and decoding process becomes more complex
Solution Approach 1:
The patent performs preliminary action by deriving temporal motion vector predictions for sub-blocks during the encoding process using motion vectors from co-located blocks in reference frames. These predictions are prepared in advance and used to construct the final motion vector, avoiding the need for more complex real-time calculations during decoding and improving overall coding efficiency.
Solution Approach 2:
The patent uses motion vectors from co-located blocks as an intermediary to derive the temporal motion vector predictions for the current block's sub-blocks. This intermediary approach allows the system to leverage existing motion information from reference frames without requiring direct complex calculations between all blocks, simplifying the overall process while maintaining efficiency.
3Manufacturing precision
If motion vectors are refined at sub-block level, then image quality is maintained at high resolutions, but the processing complexity increases
Solution Approach 1:
The patent segments the block into sub-blocks and applies temporal motion vector prediction at the sub-block level. This segmentation enables precise motion representation that maintains image quality at high resolutions by capturing local motion variations that would be lost in coarser block-based approaches, while the segmentation structure itself provides an efficient framework for processing.
Solution Approach 2:
The patent adds a temporal dimension to the motion vector prediction by deriving predictions from co-located blocks in reference frames. This temporal dimension allows the system to maintain high manufacturing precision at sub-block levels by leveraging motion information across multiple frames, reducing the processing complexity compared to purely spatial approaches at the same resolution.
Data Source
AI summary
A computing device performs a method of decoding video data by determining a co-located picture of the current coding unit; locating a spatial neighbor block of the current coding unit that corresponds to the co-located picture; determining a motion shift vector for the current coding unit from one or more motion vectors associated with the spatial neighbor block according to a predefined fixed order; and reconstructing a sub-block-based temporal motion vector for a respective sub-block of a plurality of sub-blocks in the current coding unit from a corresponding sub-block in the collocated picture based on the motion shift vector.