Granted Patent B2

US 12,621,480 · App. 18/931,329 · Granted May 5, 2026

Inventors: Takashi Hashimoto (Hyogo, JP); Takahiro Nishi (Nara, JP); Tadamasa Toma (Osaka, JP); Kiyofumi Abe (Osaka, JP); Ryuichi Kanoh (Osaka, JP)

Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICA

H04N19/513H04N19/105H04N19/176

Abstract

Claims (36)

1 . An encoder that encodes a current block using a motion vector, the encoder comprising:

a processor; and

memory, wherein

using the memory, the processor:

derives a representative motion vector indicating a representative position based on motion vector candidates included in a merge candidate list, the merge candidate list having candidates derived from motion vectors of blocks that spatially or temporally neighbor a current block;

determines, in a first reference picture for the current block, a first motion estimation region that includes the representative position indicated by the representative motion vector;

calculates first evaluation values of a plurality of candidate regions included in the first motion estimation region, the first evaluation values being differences between (i) the plurality of candidate regions and (ii) regions being along a motion trajectory of the current block in a second reference picture which is different from the first reference picture;

determines a first adjacent region included in the first motion estimation region, the first adjacent region including a first candidate region and a vicinity of the first candidate region, the first candidate region being one of the plurality of the candidate regions having a smallest first evaluation value among the plurality of candidate regions included in the first motion estimation region;

determines the motion vector of the current block using a smallest second evaluation value among a plurality of regions included in the first adjacent region;

generates reference image of the current block based on the motion vector; and

generates a bitstream which includes information indicating that a mode for motion estimation is applied.

2 . A decoder that decodes a current block using a motion vector, the decoder comprising:

a processor; and

memory, wherein

using the memory, the processor:

derives a representative motion vector indicating a representative position based on motion vector candidates included in a merge candidate list, the merge candidate list having candidates derived from motion vectors of blocks that spatially or temporally neighbor a current block;

determines, in a first reference picture for the current block, a first motion estimation region that includes the representative position indicated by the representative motion vector;

calculates first evaluation values of a plurality of candidate regions included in the first motion estimation region, the first evaluation values being differences between (i) the plurality of candidate regions and (ii) regions being along a motion trajectory of the current block in a second reference picture which is different from the first reference picture;

determines a first adjacent region included in the first motion estimation region, the first adjacent region including a first candidate region and a vicinity of the first candidate region, the first candidate region being one of the plurality of the candidate regions having a smallest first evaluation value among the plurality of candidate regions included in the first motion estimation region;

determines the motion vector of the current block using a smallest second evaluation value among a plurality of regions included in the first adjacent region; and

generates reference image of the current block based on the motion vector.

3 . An encoding method for encoding a current block using a motion vector, the encoding method comprising:

deriving a representative motion vector indicating a representative position based on motion vector candidates included in a merge candidate list, the merge candidate list having candidates derived from motion vectors of blocks that spatially or temporally neighbor a current block;

determining, in a first reference picture for the current block, a first motion estimation region that includes the representative position indicated by the representative motion vector;

calculating first evaluation values of a plurality of candidate regions included in the first motion estimation region, the first evaluation values being differences between (i) the plurality of candidate regions and (ii) regions being along a motion trajectory of the current block in a second reference picture which is different from the first reference picture;

determining a first adjacent region included in the first motion estimation region, the first adjacent region including a first candidate region and a vicinity of the first candidate region, the first candidate region being one of the plurality of the candidate regions having a smallest first evaluation value among the plurality of candidate regions included in the first motion estimation region;

determining the motion vector of the current block using a smallest second evaluation value among a plurality of regions included in the first adjacent region;

generating reference image of the current block based on the motion vector; and

generating a bitstream which includes information indicating that a mode for motion estimation is applied.

4 . A decoding method for decoding a current block using a motion vector, the decoding method comprising:

deriving a representative motion vector indicating a representative position based on motion vector candidates included in a merge candidate list, the merge candidate list having candidates derived from motion vectors of blocks that spatially or temporally neighbor a current block;

determining, in a first reference picture for the current block, a first motion estimation region that includes the representative position indicated by the representative motion vector;

calculating first evaluation values of a plurality of candidate regions included in the first motion estimation region, the first evaluation values being differences between (i) the plurality of candidate regions and (ii) regions being along a motion trajectory of the current block in a second reference picture which is different from the first reference picture;

determining a first adjacent region included in the first motion estimation region, the first adjacent region including a first candidate region and a vicinity of the first candidate region, the first candidate region being one of the plurality of the candidate regions having a smallest first evaluation value among the plurality of candidate regions included in the first motion estimation region;

determining the motion vector of the current block using a smallest second evaluation value among a plurality of regions included in the first adjacent region; and

generating reference image of the current block based on the motion vector.

Priority Claims (1)

Continuity (8)

References Cited (21)