Motion Estimation Update Tables for Low-Cost Sub-Pixel Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion estimation methods face challenges in improving conversion performance and sub-pixel estimation granularity while maintaining computational efficiency, as the number of updates required increases, leading to high computational burdens.
Innovation Solution
Distributing updates for predictor vectors across multiple tables, allowing only a single table's updates to be applied to generate candidate vectors, thereby reducing the computational burden without compromising estimation quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of updates applied to predictor vector is increased, then conversion performance and sub-pixel estimation granularity are improved, but computational burden increases
Solution Approach 1:
The set of updates is divided into multiple tables, where each table contains a subset of updates. Only one table is applied to the predictor vector at a time, segmenting the computational work into manageable portions that can be processed efficiently while maintaining access to a large total number of updates for high precision estimation.
2Reliability
If the number of updates applied to predictor vector is increased, then conversion performance is improved, but computational burden increases
Solution Approach 1:
The updates are organized into multiple tables that can be selectively applied. This segmentation allows the system to achieve high conversion performance by having access to a large total number of updates across all tables, while the computational burden at any given time is limited to processing only one table's worth of updates.
3Manufacturing precision
If a large number of updates are generated and applied to predictor vector, then candidate vector quality is improved, but processing time increases
Solution Approach 1:
By dividing updates into multiple tables and applying only one table at a time, the system maintains high candidate vector quality through access to a large total number of updates, while reducing processing time by limiting the number of updates that must be processed in each iteration.
Solution Approach 2:
The method employs periodic application of different update tables in recursive iterations. Each iteration applies updates from a single table periodically, allowing the system to cycle through multiple tables across different iterations, thereby achieving high quality results without excessive processing time in any single pass.
Data Source
AI summary
A method for motion estimation with respect to a sequence of images. Respective updates are generated based on predictor vectors. The updates are distributed over a plurality of tables, which are organized in a plurality of table sets. Only a single table of updates, of a respective selected table set, is applied to a given predictor vector to generate a limited set of candidate vectors only. For a subsequent predictor vector, a further single table of the plurality of tables out of the plurality of table sets is applied.