US 11,985,320 B2
Early termination for optical flow refinement
Semih Esenlik, Munich (DE); Sriram Sethuraman, Bangalore (IN); Jeeva Raj A, Rasipuram (IN); and Sagar Kotecha, Mangal Karyalaya (IN)
Assigned to HUAWEI TECHNOLOGIES CO., LTD., Guangdong (CN)
Filed by HUAWEI TECHNOLOGIES CO., LTD., Guangdong (CN)
Filed on Aug. 20, 2021, as Appl. No. 17/407,617.
Application 17/407,617 is a continuation of application No. PCT/CN2020/076178, filed on Feb. 21, 2020.
Claims priority of application No. 201931007114 (IN), filed on Feb. 22, 2019.
Prior Publication US 2021/0392334 A1, Dec. 16, 2021
Int. Cl. H04N 19/132 (2014.01); H04N 19/117 (2014.01); H04N 19/159 (2014.01); H04N 19/176 (2014.01); H04N 19/82 (2014.01)
CPC H04N 19/132 (2014.11) [H04N 19/117 (2014.11); H04N 19/159 (2014.11); H04N 19/176 (2014.11); H04N 19/82 (2014.11)] 25 Claims
OG exemplary drawing
 
1. A method of video coding implemented in a decoding device or an encoding device, the method comprising:
obtaining a pair of initial motion vectors for a current block;
obtaining two first prediction blocks for the current block based on the pair of initial motion vectors;
calculating a first matching cost according to the two first prediction blocks;
obtaining a pair of refined motion vectors and a second matching cost corresponding to the pair of refined motion vectors;
obtaining two second prediction blocks for the current block according to the refined motion vectors:
determining whether an optical flow refinement process should be performed, according to at least one preset condition, the at least one preset condition comprising a condition of whether the second matching cost is equal to or larger than a threshold value; and
performing an optical flow refinement process based on the two second prediction blocks for obtaining a final inter prediction for the current block, in response to determining that the optical flow refinement process should be performed;
when it is determined that the optical flow refinement process should not be performed, the final inter prediction is obtained by a weighted sum of the two second prediction blocks.