Motion Vector Predictor Scaling for Inter-Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image encoding/decoding technologies face challenges in accurately predicting pixel values for high-resolution and high-definition images, particularly in inter-prediction methods, where motion vector prediction is crucial for efficient encoding and decoding.
Innovation Solution
The method generates a motion vector predictor for a target block based on the motion vector of a candidate block in a co-located picture, using scaling coefficients and selecting appropriate motion vectors to improve prediction accuracy, especially in Advanced Motion Vector Prediction, merge, and skip modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If inter-prediction technology is used to predict pixel values using temporally previous pictures, then encoding efficiency is improved, but prediction accuracy deteriorates for high-resolution and high-definition images
Solution Approach 1:
The patent applies parameter changes by introducing scaling coefficients (temporal scaling and spatial scaling) to adjust motion vectors according to different picture resolutions and temporal distances. This allows the motion vector prediction to adapt to high-resolution images by scaling the motion information from candidate blocks, thereby improving prediction accuracy while maintaining encoding efficiency.
Solution Approach 2:
The patent uses motion vector predictors as intermediaries between candidate blocks and target blocks. By generating motion vector predictors from multiple candidate blocks (including co-located picture candidates and spatial neighbor candidates) and combining them with scaling coefficients, the system creates an intermediate representation that improves prediction accuracy for high-resolution images.
2Device complexity
If motion vector prediction is performed using candidate blocks, then encoding complexity is reduced, but prediction precision deteriorates
Solution Approach 1:
The patent merges multiple motion vector prediction approaches by combining co-located picture candidate blocks with spatial neighbor candidate blocks. The motion vector predictor is generated by combining motion vectors from different candidate blocks with appropriate scaling coefficients, achieving higher prediction precision without significantly increasing encoding complexity.
Solution Approach 2:
The patent introduces scaling coefficients as parameters to adjust motion vectors based on temporal distance and spatial resolution. This parameter change allows the system to maintain low encoding complexity while improving motion vector prediction precision by adapting to different picture resolutions and temporal relationships.
3Measurement precision
If scaling coefficients are applied to motion vectors, then prediction accuracy is improved, but computational load increases
Solution Approach 1:
The patent applies local quality by selectively applying scaling coefficients only where needed - specifically for co-located picture candidate blocks where temporal distance and resolution differences exist. Spatial neighbor candidates that do not require scaling are processed normally, thereby improving prediction accuracy where necessary while minimizing the increase in computational load.
Data Source
AI summary
Disclosed herein are a decoding method and apparatus and an encoding method and apparatus that perform inter-prediction using a motion vector predictor. For a candidate block in a col picture, a scaled motion vector is generated based on a motion vector of the candidate block. When the scaled motion vector indicates a target block, a motion vector predictor of the target block is generated based on the motion vector of the candidate block. The motion vector predictor is used to derive the motion vector of the target block in a specific inter-prediction mode such as a merge mode and an AMVP mode.