Motion Vector Prediction for Video Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video codecs face inefficiencies in encoding and decoding high-resolution video content due to limitations in motion vector prediction methods, particularly when dealing with blocks of predetermined sizes and reference images classified as short-term or long-term reference images.
Innovation Solution
A method and apparatus for determining motion vectors through motion vector prediction, where candidate blocks are identified based on their reference images' classification as long-term or short-term reference images, allowing for optimized motion vector selection and prediction, thereby improving prediction efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion vector prediction is performed using all candidate blocks regardless of reference image type, then prediction accuracy may be maintained, but encoding complexity and processing time increase
Solution Approach 1:
The patent segments the candidate blocks into two distinct groups based on reference image type: short-term reference image blocks and long-term reference image blocks. This segmentation allows the encoder to apply different processing strategies to each group, specifically omitting certain motion vector operations for long-term reference blocks to reduce complexity while maintaining overall prediction accuracy through selective processing of the more critical short-term reference blocks
Solution Approach 2:
The patent extracts and identifies long-term reference image blocks from the set of candidate blocks and excludes them from certain motion vector prediction operations. By taking out these specific blocks that contribute less to prediction accuracy, the encoder reduces processing complexity without significantly compromising the overall prediction quality, as the essential motion information is captured through short-term reference blocks
2Loss of information
If motion vector operations are performed on all candidate blocks, then comprehensive motion information is captured, but processing time and computational load increase
Solution Approach 1:
The patent applies partial action by performing motion vector operations only on candidate blocks that reference short-term reference images, while omitting similar operations for long-term reference image blocks. This selective approach captures the most critical motion information needed for accurate prediction while avoiding unnecessary computational expenditure on blocks that contribute less significantly to prediction accuracy, thereby reducing processing time without excessive loss of motion information
3Reliability
If candidate blocks from long-term reference images are processed with full motion vector operations, then prediction accuracy is maintained, but encoding efficiency decreases
Solution Approach 1:
The patent applies local quality by differentiating the processing quality applied to different types of candidate blocks. Short-term reference image blocks receive full motion vector prediction processing to ensure high accuracy, while long-term reference image blocks receive reduced or omitted processing. This localized quality adjustment optimizes encoding efficiency by concentrating computational resources on blocks that provide the most significant contribution to prediction accuracy
Data Source
AI summary
Provided are a motion vector predicting method and an apparatus for encoding and decoding a video. The motion vector prediction method includes: determining, from neighboring blocks of the current block, a plurality of candidate blocks that are referred to so as to predict a motion vector of a current block; determining a candidate motion vector of a first candidate block among the determined plurality of candidate blocks, based on whether a reference image of the first candidate block and a reference image of the current block are long-term reference images; and determining the motion vector of the current block by using a candidate motion vector list including the determined candidate motion vector of the first candidate block and candidate motion vectors from remaining candidate blocks among the determined plurality of candidate blocks.