US 11,895,298 B2
Encoder, decoder, encoding method, and decoding method
Kiyofumi Abe, Osaka (JP); Takahiro Nishi, Nara (JP); Tadamasa Toma, Osaka (JP); Ryuichi Kanoh, Osaka (JP); Chong Soon Lim, Singapore (SG); Ru Ling Liao, Singapore (SG); Hai Wei Sun, Singapore (SG); Sughosh Pavan Shashidhar, Singapore (SG); Han Boon Teo, Singapore (SG); and Jing Ya Li, Singapore (SG)
Assigned to Panasonic Intellectual Property Corporation of America, Torrance, CA (US)
Filed by Panasonic Intellectual Property Corporation of America, Torrance, CA (US)
Filed on Apr. 19, 2022, as Appl. No. 17/724,178.
Application 17/724,178 is a continuation of application No. 16/942,630, filed on Jul. 29, 2020, granted, now 11,350,092.
Application 16/942,630 is a continuation of application No. PCT/JP2019/003065, filed on Jan. 30, 2019.
Claims priority of provisional application 62/623,834, filed on Jan. 30, 2018.
Prior Publication US 2022/0239913 A1, Jul. 28, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. H04N 19/52 (2014.01); H04N 19/119 (2014.01); H04N 19/137 (2014.01); H04N 19/176 (2014.01)
CPC H04N 19/119 (2014.11) [H04N 19/137 (2014.11); H04N 19/176 (2014.11); H04N 19/52 (2014.11)] 3 Claims
OG exemplary drawing
 
1. An encoder, comprising:
circuitry; and
memory, wherein
using the memory, the circuitry:
determines to perform first inter prediction or second inter prediction based on information on inter prediction, the first inter prediction including a bi-directional optical flow (BIO) process, the second inter prediction including a motion compensation process, the second inter prediction not including the BIO process; and
encodes an image block based on the determined inter prediction process, wherein,
the image block is encoded in a bitstream,
when the BIO process is determined to be performed, the circuitry (i) obtains first prediction images for the image block, (ii) obtains gradient images for the first prediction images, and (iii) generates a prediction image for the image block using the first prediction images and the gradient images, and
when the motion compensation process is determined to be performed, the circuitry (i) determines a plurality of partitions in the image block, (ii) obtains a second prediction image for each of the plurality of partitions, and (iii) generates a prediction image for the image block using the second prediction images.