Adaptive Motion Vector Search Ordering for Video Encoder Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern video encoders face high computational complexity in motion estimation due to the evaluation of all possible block matching candidates, leading to unnecessary cost function evaluations and reduced accuracy when using algorithms that do not preserve optimality.
Innovation Solution
Adaptive search ordering for rate-constrained successive elimination methods that only perform necessary cost function evaluations by sorting candidate motion vectors based on approximate block similarity metrics and using an early termination threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exhaustive search algorithm is used to evaluate all block matching candidates, then motion estimation accuracy is improved, but computational complexity increases significantly
Solution Approach 1:
The patent applies preliminary action by computing approximate block similarity metrics (ADS) for all candidate motion vectors before the main cost function evaluation. This pre-computation creates a sorted list of candidates based on their approximate similarity, allowing the algorithm to evaluate candidates in order of likelihood without computing full cost functions for all candidates, thus reducing computational complexity while maintaining accuracy.
Solution Approach 2:
The patent segments the motion estimation process into two distinct phases: (1) computing approximate block similarity metrics for all candidates and sorting them, and (2) sequentially evaluating cost functions only for candidates above the early termination threshold. This segmentation allows the algorithm to avoid unnecessary full cost function evaluations for unlikely candidates.
2Productivity
If algorithms that do not preserve optimality are used to reduce computational complexity, then processing speed is improved, but motion estimation accuracy deteriorates
Solution Approach 1:
The patent changes the parameter of block similarity measurement by introducing approximate block similarity metrics (ADS) as a preliminary filtering criterion. This parameter change allows the algorithm to quickly eliminate unlikely candidates using a simplified metric before applying the full cost function, thereby improving processing speed without sacrificing the accuracy of the final selection.
Solution Approach 2:
The patent introduces an intermediary element - the approximate block similarity metric (ADS) - that serves as a mediator between the exhaustive search and the simplified search. This intermediary allows the algorithm to maintain optimality by providing a reliable preliminary assessment that guides the selection of candidates for full evaluation, ensuring that the true optimal candidate is not missed while reducing computational burden.
3Measurement precision
If more block matching candidates are evaluated, then motion estimation accuracy is improved, but the number of necessary cost function evaluations increases
Solution Approach 1:
The patent applies partial action by evaluating cost functions only for a subset of candidates - specifically those whose approximate block similarity metric exceeds the early termination threshold. This partial evaluation approach is sufficient to find the optimal motion vector without performing excessive evaluations on candidates that are unlikely to be optimal, thus reducing the number of costly full evaluations while maintaining accuracy.
Data Source
AI summary
Methods and systems for determining motion vectors in a motion estimation process of a video encoder are provided. Embodiments of the present invention provide a solution for the problem of unnecessary cost function evaluations, found when combining the successive elimination method with a predetermined list of candidate motion vectors search ordering. In order to solve this problem, embodiments of the present invention provide an adaptive scan ordering of block matching candidates within a search area. Advantageously, embodiments of the present invention will only evaluate necessary cost functions, without impacting rate-distortion.