US Patent Application 18313202. MOTION VECTOR REFINEMENT FOR MULTI-REFERENCE PREDICTION simplified abstract
Contents
MOTION VECTOR REFINEMENT FOR MULTI-REFERENCE PREDICTION
Organization Name
Inventor(s)
Anand Meher Kotra of Munich (DE)
MOTION VECTOR REFINEMENT FOR MULTI-REFERENCE PREDICTION - A simplified explanation of the abstract
This abstract first appeared for US patent application 18313202 titled 'MOTION VECTOR REFINEMENT FOR MULTI-REFERENCE PREDICTION
Simplified Explanation
The patent application is about improving the performance of motion vector refinement in multi-reference inter-prediction.
- Two or more reference pictures are selected for the refinement process.
- An initial estimate of a motion vector to one of the reference pictures is made.
- A search space is created in the selected reference image based on the initial estimate.
- Template matching is used to refine the first motion vector.
- The second motion vector to another reference picture is calculated using its initial estimate, the initial estimate of the first motion vector, and the refined first motion vector.
Original Abstract Submitted
The present invention relates to the performance of motion vector refinement in a search space for multi-reference inter-prediction. Two or more reference pictures are selected, one of those used for motion vector refinement. Based on an initial estimate of a motion vector to the reference picture for motion vector refinement, a search space in this reference image is constructed. Using template matching, the first motion vector is refined. The second motion vector to another reference picture is calculated using its initial estimate, the initial estimate of the first motion vector and the refined first motion vector.