Inter-frame Prediction Method Reducing Motion Estimation Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of selecting a target reference frame in inter-frame prediction processes hampers the efficiency of video encoding due to the high processing burden of motion estimation, necessitating a reduction in the number of motion estimation searches.
Innovation Solution
An inter-frame prediction method that determines advanced motion vector predictions (AMVPs) for adjacent reference encoded blocks, selects a target AMVP, and compares it with a predetermined reference frame to perform motion estimation only when different, thereby reducing the number of motion estimation searches and selecting the target reference frame based on encoding costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion estimation is performed on all reference frames to select the optimal target reference frame, then the accuracy of prediction encoding is improved, but the processing complexity and encoding time increase significantly
Solution Approach 1:
The patent performs preliminary action by calculating AMVP values for all reference frames before the actual motion estimation process. This preliminary calculation identifies candidate reference frames with smallest AMVP values, so that full motion estimation is only performed on these candidates rather than all reference frames, thereby reducing processing complexity while maintaining prediction accuracy
Solution Approach 2:
The patent segments the reference frame selection process into two stages: first calculating AMVP values for all reference frames to identify candidates, then performing motion estimation only on candidate frames. This segmentation divides the complex task into manageable parts, reducing overall processing complexity while preserving accuracy
2Manufacturing precision
If motion estimation is performed on all reference frames to select the optimal target reference frame, then the quality of video encoding is improved, but the encoding time increases
Solution Approach 1:
The patent performs preliminary calculation of AMVP values for all reference frames before motion estimation. This preliminary action identifies candidate reference frames that are most likely to yield optimal encoding quality, allowing the system to skip full motion estimation on non-candidate frames and thus reduce encoding time while maintaining quality
Solution Approach 2:
The patent performs partial motion estimation only on candidate reference frames identified through AMVP calculation, rather than performing full motion estimation on all reference frames. This partial action is sufficient to achieve optimal encoding quality while significantly reducing encoding time
3Productivity
If the number of motion estimation searches is reduced to decrease processing complexity, then the encoding efficiency is improved, but the risk of selecting a suboptimal reference frame increases
Solution Approach 1:
The patent uses AMVP calculation as a feedback mechanism to evaluate all reference frames and identify candidates with smallest AMVP values. This feedback guides the subsequent motion estimation process by indicating which reference frames are most promising, ensuring that reduced searches still target the optimal frames and maintaining selection reliability
Solution Approach 2:
The patent introduces AMVP calculation as an intermediary step between having all reference frames available and performing motion estimation. This intermediary process filters and ranks reference frames, allowing the system to reduce the number of motion estimation searches while maintaining reliability by always selecting from the best candidates identified through the intermediary AMVP evaluation
Data Source
AI summary
Embodiments of the present disclosure provide an inter-frame prediction method performed at a computing device. The method includes: determining at least one reference encoded block spatially adjacent to a to-be-processed prediction unit of a to-be-processed encoded block; for each preset reference frame, separately determining a corresponding AMVP of the to-be-processed prediction unit according to reference information of each reference encoded block; determining a target AMVP under the reference frames, and using a reference frame corresponding to the target AMVP as a candidate reference frame; when the candidate reference frame is different from the first reference frame, performing motion estimation on the to-be-processed prediction unit separately and determining a target reference frame from the candidate reference frame and the first reference frame according to their corresponding encoding costs obtained through the motion estimation.