US 12,120,353 B2
Image decoding device, method, and program with prediction refinement optical flow
Kyohei Unno, Saitama (JP); Kei Kawamura, Saitama (JP); and Sei Naito, Saitama (JP)
Assigned to KDDI CORPORATION, Tokyo (JP)
Filed by KDDI CORPORATION, Tokyo (JP)
Filed on Feb. 17, 2022, as Appl. No. 17/674,164.
Application 17/674,164 is a continuation of application No. PCT/JP2020/031562, filed on Aug. 21, 2020.
Claims priority of application No. 2019-172127 (JP), filed on Sep. 20, 2019.
Prior Publication US 2022/0174325 A1, Jun. 2, 2022
Int. Cl. H04N 19/70 (2014.01); H04N 19/139 (2014.01); H04N 19/172 (2014.01); H04N 19/174 (2014.01); H04N 19/46 (2014.01); H04N 19/50 (2014.01); H04N 19/51 (2014.01); H04N 19/82 (2014.01)
CPC H04N 19/70 (2014.11) [H04N 19/139 (2014.11); H04N 19/172 (2014.11); H04N 19/174 (2014.11); H04N 19/46 (2014.11); H04N 19/50 (2014.11); H04N 19/51 (2014.11); H04N 19/82 (2014.11)] 3 Claims
OG exemplary drawing
 
1. An image decoding device comprising a circuit, wherein the circuit:
decodes a first flag which is included in a sequence parameter set and controls enabling/disabling of PROF (Prediction Refinement with Optical Flow) processing from encoded data;
decodes a second flag indicating whether or not a flag indicating enabling/disabling of the PROF processing is included in a low order header of a picture parameter set from the sequence parameter set additionally in a case where the first flag indicates enabling of the PROF processing in units of sequences;
decodes a third flag which is included in a low order header of the picture parameter set and indicates enabling/disabling of the PROF processing from the encoded data in a case where the second flag indicates that the flag indicating enabling/disabling of the PROF processing is included in the low order header of a picture parameter set;
determines non-application conditions of the PROF processing for each block;
determines not to apply the PROF processing to a block that meets at least one of the non-application conditions of the PROF processing and generate a prediction signal without performing the PROF processing; and
calculates “diffMV” corresponding to each pixel position in each sub-block to which a block that doesn't meet any of the non-application conditions of the PROF processing is divided and performs correction of the prediction signal by executing the PROF processing with the “diffMV”,
the non-application conditions of the PROF processing include a condition that the first flag indicates disabling of the PROF processing in units of sequences,
the non-application conditions of the PROF processing include a condition that the third flag indicates disabling of the PROF processing, and
the non-application conditions of the PROF processing include a condition that “fallbackMode” is applied to a block.