Decoder-Side FRUC Refinement for Accurate Video Motion Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding technologies, particularly in HEVC, face inefficiencies in the refinement of motion vectors at sub-pixel levels, which affect coding efficiency and bitrate costs.
Innovation Solution
The method involves a decoder-side motion vector derivation process that refines motion vectors at sub-pixel positions using multiple search steps and patterns, including diamond and cross patterns, and adapts these patterns based on the signal characteristics and edge directions within the block, to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vector refinement is performed at conventional sub-pixel resolutions (1/4 or 1/8 pixel), then the decoding complexity is reduced, but the coding efficiency and prediction accuracy deteriorate
Solution Approach 1:
The refinement process is divided into multiple search steps, each operating at a different sub-pixel resolution level. The first search step uses 1/4 pixel precision, the second uses 1/8 pixel precision, and the third uses 1/16 pixel precision. This segmentation allows the system to achieve high precision without processing every possible sub-pixel position simultaneously, thereby reducing computational complexity while maintaining accuracy.
Solution Approach 2:
The patent dynamically adjusts the search pattern and resolution level based on the characteristics of the current block. The encoder evaluates whether to perform refinement at 1/16 pixel precision or skip to coarser resolutions based on block properties, creating a dynamic adaptation mechanism that balances precision requirements with computational complexity for each specific case.
2Productivity
If multiple search steps with increasing sub-pixel resolution are implemented, then coding efficiency is improved, but processing time and computational load increase
Solution Approach 1:
The refinement process uses periodic search patterns (diamond pattern followed by cross patterns) that systematically sample sub-pixel positions in a repeating sequence. This periodic sampling allows the system to efficiently explore the search space without visiting every possible position, reducing processing time while maintaining coding efficiency through structured evaluation of motion vector candidates.
Solution Approach 2:
The patent changes the resolution parameter across different search steps, transitioning from 1/4 pixel to 1/8 pixel to 1/16 pixel precision. By adjusting this parameter dynamically, the system can achieve higher coding efficiency when needed while avoiding the full computational burden of always operating at the highest precision level, thus optimizing the trade-off between processing time and coding performance.
3Measurement precision
If adaptive search patterns based on block characteristics are used, then prediction accuracy is enhanced, but the complexity of determining search patterns increases
Solution Approach 1:
The encoder performs preliminary evaluation of block characteristics before initiating the refinement process. It determines in advance whether to apply 1/16 pixel precision refinement or use coarser resolutions based on pre-analysis of the block's properties. This preliminary action simplifies the overall complexity by avoiding unnecessary complex processing for blocks that do not require high precision, while still achieving enhanced accuracy when needed.
Data Source
AI summary
A method and a device for encoding or decoding video data. It concerns more particularly the encoding according to a particular encoding mode using a decoder side motion vector derivation mode referenced as frame-rate up conversion mode or FRUC mode. In FRUC merge mode, the derivation process comprises a refinement step to increase the accuracy of the obtained motion vector at the sub-pixel level. This process involves the evaluation of different sub-pixel position around the obtained motion vector according to different patterns. The present invention has been devised to improve the known refinement step. It aims at improving the coding efficiency by considering the characteristics of the matching type and/or the signal inside the templates.