Weighted External Motion Vectors for Video Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video pipeline systems face inefficiencies and sub-optimal motion vector fields due to the inability to reuse motion vectors effectively across different blocks and stages, primarily because of differing requirements and risks associated with introducing motion vectors from external sources.
Innovation Solution
The proposed method involves subdividing video frames or fields into blocks and assigning candidate motion vectors, including those derived from external source vector fields, with each candidate assigned a weighting value to bias towards or away from selection, thereby improving the efficiency and quality of motion estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vectors from external sources are introduced into the candidate set, then the quality and accuracy of motion estimation is improved, but the complexity of the system increases due to weighting mechanisms and external source integration
Solution Approach 1:
The patent introduces a weighting mechanism as an intermediary between external motion vector sources and the motion estimation algorithm. Each external motion vector candidate is assigned a weight that reflects its reliability, allowing the system to selectively trust certain external sources while maintaining the ability to reject unreliable ones. This mediator layer enables quality improvement without requiring complete integration of all external sources, thus managing complexity.
Solution Approach 2:
The system changes the parameter of motion vector candidates by assigning weighting values that modify their influence in the cost function. Instead of treating all candidates equally, the weights dynamically adjust the effective strength of each candidate based on external source reliability, motion complexity, and matching quality. This parameter change allows the system to adapt to different external source qualities without restructuring the entire motion estimation framework.
2Measurement precision
If a larger number of motion vector candidates are tested, then the quality of motion estimation is improved, but the computational resources and time required increase
Solution Approach 1:
The patent applies partial action by testing only a subset of external motion vector candidates rather than exhaustively evaluating all possible candidates. The weighting mechanism allows the system to focus computational effort on high-weight candidates that are most likely to provide accurate motion vectors. This partial evaluation approach maintains quality by concentrating resources on promising candidates while avoiding the excessive computational cost of testing all candidates.
Solution Approach 2:
The system performs preliminary action by pre-calculating and assigning weights to external motion vector candidates before the main motion estimation process. External sources provide motion vector candidates with associated weightings in advance, allowing the motion estimation algorithm to quickly evaluate and prioritize candidates without performing extensive preliminary analysis during the time-critical encoding phase. This preliminary weighting reduces the computational burden during actual motion estimation.
3Productivity
If motion vectors from external sources are reused, then computational efficiency is improved, but reliability decreases due to mismatched requirements between different motion estimation stages
Solution Approach 1:
The patent introduces dynamics by making the weighting of external motion vectors adaptive rather than static. The weights are adjusted based on the specific requirements of each motion estimation stage, the quality of match between external candidates and current block content, and the reliability of the external source for that particular scenario. This dynamic weighting allows the system to reliably reuse external vectors when appropriate while automatically reducing reliance on them when requirements mismatch, thus maintaining both efficiency and reliability.
Solution Approach 2:
The system implements feedback mechanisms where the performance of externally sourced motion vectors is evaluated and used to adjust future weighting decisions. If external motion vectors consistently provide accurate results for certain block types or motion patterns, their weights are increased for similar future cases. Conversely, if external vectors prove unreliable for certain conditions, their weights are reduced. This feedback loop enables reliable reuse of external vectors by learning from past performance while adapting to different requirements across various motion estimation stages.
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
A method and apparatus for motion estimation in a sequence of video images is provided. Each field or frame in a sequence of video images is sub-divided into a plurality of blocks. Each block in each video field or frame has assigned to it a set of candidate motion vectors. The vector which produces a best match to a block in a previous field or frame, from the set of candidate motion vectors, is assigned to that block thus forming the motion vector field for the current video field or frame using the selected vector. The set of candidate motion vectors assigned to a block include one or more candidate vectors derived from an external source vector field and each such vector from an external source vector field is assigned a weighting that biases towards or away from the selection of that vector.