Granted Patent B2

US 12,739,398 · App. 19/476,412 · Granted Sep 15, 2026

Inventors: Yunzhuo Liu (Hangzhou, CN); Fangdong Chen (Hangzhou, CN); Li Wang (Hangzhou, CN); Xiaoyang Wu (Hangzhou, CN)

Assignee: HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO., LTD.

H04N19/14H04N19/176H04N19/196

Abstract

Claims (65)

1 . A decoding method, comprising:

decoding a bitstream corresponding to a current picture block to obtain a coefficient hyperparameter feature corresponding to the current picture block;

determining a probability distribution parameter based on the coefficient hyperparameter feature;

based on the probability distribution parameter, decoding a bitstream corresponding to the current picture block to obtain an initial reconstructed feature of the current picture block; and

determining a target reconstructed picture block corresponding to the current picture block based on the initial reconstructed feature,

wherein determining the target reconstructed picture block corresponding to the current picture block based on the initial reconstructed feature comprises:

inputting the initial reconstructed feature into a synthesis transform network to obtain an initial reconstructed picture block corresponding to the current picture block; and

based on the probability distribution parameter and a picture domain enhancement parameter corresponding to the current picture block, performing picture adaptive edge enhancement on the initial reconstructed picture block to obtain the target reconstructed picture block corresponding to the current picture block,

wherein the picture domain enhancement parameter is obtained by decoding a bitstream corresponding to the current picture block.

2 . The method according to claim 1 , wherein determining the target reconstructed picture block corresponding to the current picture block based on the initial reconstructed feature comprises:

inputting the initial reconstructed feature into a synthesis transform network to obtain the target reconstructed picture block corresponding to the current picture block.

3 . The method according to claim 1 , wherein the picture domain enhancement parameter comprises picture enhancement segment magnitude values and picture enhancement segment thresholds, wherein the picture enhancement segment thresholds form picture enhancement threshold intervals, and the picture enhancement threshold intervals respectively correspond to the picture enhancement segment magnitude values.

4 . The method according to claim 3 , wherein based on the probability distribution parameter and the picture domain enhancement parameter corresponding to the current picture block, performing picture adaptive edge enhancement on the initial reconstructed picture block to obtain the target reconstructed picture block corresponding to the current picture block comprising:

obtaining a target probability distribution channel map based on the probability distribution parameter;

when the target probability distribution channel map comprises a plurality of probability distribution values, for each of the plurality of probability distribution values, determining the picture enhancement segment magnitude value corresponding to the probability distribution value based on the picture enhancement threshold interval corresponding to the probability distribution value; and

based on the picture enhancement segment magnitude value corresponding to each of the plurality of probability distribution values, performing picture adaptive edge enhancement on the initial reconstructed picture block to obtain the target reconstructed picture block corresponding to the current picture block.

5 . The method according to claim 4 , wherein obtaining the target probability distribution channel map based on the probability distribution parameter comprises:

when the probability distribution parameter comprises an important probability distribution channel map and a non-important probability distribution channel map, upsampling the important probability distribution channel map to obtain the target probability distribution channel map;

wherein a size of the target probability distribution channel map is same as a size of the initial reconstructed picture block.

6 . The method according to claim 4 , wherein based on the picture enhancement segment magnitude value corresponding to each of the plurality of probability distribution values, performing picture adaptive edge enhancement on the initial reconstructed picture block to obtain the target reconstructed picture block corresponding to the current picture block comprises:

generating a high-frequency detail picture based on the initial reconstructed picture block;

for each feature value in the high-frequency detail picture, based on the picture enhancement segment magnitude value corresponding to the probability distribution value corresponding to the feature value, performing edge enhancement on the feature value, to obtain a picture enhanced feature value; and

determining the target reconstructed picture block based on the picture enhanced feature value corresponding to each feature value in the high-frequency detail picture.

7 . The method according to claim 1 , wherein based on the probability distribution parameter and the picture domain enhancement parameter corresponding to the current picture block, performing picture adaptive edge enhancement on the initial reconstructed picture block to obtain the target reconstructed picture block corresponding to the current picture block comprising:

based on the picture domain enhancement parameter and the probability distribution parameter, performing picture adaptive edge enhancement on an initial reconstructed picture block corresponding to a luma component of the current picture block, to obtain a target reconstructed picture block corresponding to the luma component.

8 . The method according to claim 5 , wherein the initial reconstructed feature comprises a plurality of feature channel maps, the probability distribution parameter comprises a plurality of probability distribution channel maps, the plurality of probability distribution channel maps respectively correspond to the plurality of feature channel maps, and the method further comprises:

decoding a bitstream corresponding to the current picture block to obtain an important channel identifier;

based on the important channel identifier, selecting a feature channel map corresponding to the important channel identifier from the plurality of feature channel maps as an important feature channel map, and selecting remaining feature channel maps as non-important feature channel maps; and

selecting a probability distribution channel map corresponding to the important feature channel map as the important probability distribution channel map, and selecting probability distribution channel maps corresponding to the non-important feature channel maps as the non-important probability distribution channel maps.

9 . The method according to claim 5 , wherein the initial reconstructed feature comprises a plurality of feature channel maps, the probability distribution parameter comprises a plurality of probability distribution channel maps, the plurality of probability distribution channel maps respectively correspond to the plurality of feature channel maps, and the method further comprises:

for each of the plurality of feature channel maps, based on feature values in the feature channel map and probability distribution values in the probability distribution channel map corresponding to the feature channel map, determining a number of consumed bits for the feature channel map;

based on the number of consumed bits for each of the plurality of feature channel maps, selecting an important feature channel map from the plurality of feature channel maps, and selecting remaining feature channel maps as non-important feature channel maps; and

selecting a probability distribution channel map corresponding to the important feature channel map as the important probability distribution channel map, and selecting probability distribution channel maps corresponding to the non-important feature channel maps as the non-important probability distribution channel maps.

10 . The method according to claim 8 , wherein the probability distribution parameter and the coefficient hyperparameter feature corresponding to the current picture block are obtained by decoding a first bitstream corresponding to the current picture block;

by decoding a second bitstream corresponding to the current picture block, obtaining the initial reconstructed feature corresponding to the current picture block; and

by decoding a third bitstream corresponding to the current picture block, obtaining the important channel identifier;

wherein the first bitstream, the second bitstream, and the third bitstream are bitstreams encoding different information.

11 . The method according to claim 10 , wherein the third bitstream is a header information bitstream.

12 . An encoding method, comprising:

encoding a coefficient hyperparameter feature corresponding to a current picture block to obtain a first bitstream corresponding to the current picture block;

determining a probability distribution parameter based on the coefficient hyperparameter feature;

encoding an initial picture feature corresponding to the current picture block based on the probability distribution parameter, to obtain a second bitstream corresponding to the current picture block; and

encoding an important channel identifier to obtain a third bitstream corresponding to the current picture block,

wherein encoding the important channel identifier to obtain the third bitstream corresponding to the current picture block comprises:

encoding a picture domain enhancement parameter to obtain the third bitstream,

wherein the picture domain enhancement parameter is configured to perform, by a decoder, picture adaptive edge enhancement on an initial reconstructed picture block based on the picture domain enhancement parameter and the probability distribution parameter to obtain a target reconstructed picture block corresponding to the current picture block.

13 . A decoding device, comprising:

one or more processors; and

one or more machine-readable storage media, wherein the one or more machine-readable storage media store machine executable instructions that can be executed by the one or more processors,

wherein the one or more processors are configured to execute the machine executable instructions to perform operations comprising:

decoding a bitstream corresponding to a current picture block to obtain a coefficient hyperparameter feature corresponding to the current picture block;

determining a probability distribution parameter based on the coefficient hyperparameter feature;

based on the probability distribution parameter, decoding a bitstream corresponding to the current picture block to obtain an initial reconstructed feature of the current picture block; and

determining a target reconstructed picture block corresponding to the current picture block based on the initial reconstructed feature,

wherein determining the target reconstructed picture block corresponding to the current picture block based on the initial reconstructed feature comprises:

inputting the initial reconstructed feature into a synthesis transform network to obtain an initial reconstructed picture block corresponding to the current picture block; and

based on the probability distribution parameter and a picture domain enhancement parameter corresponding to the current picture block, performing picture adaptive edge enhancement on the initial reconstructed picture block to obtain the target reconstructed picture block corresponding to the current picture block,

wherein the picture domain enhancement parameter is obtained by decoding a bitstream corresponding to the current picture block.

14 . An encoder device, comprising:

one or more processors; and

one or more machine-readable storage media,

wherein the one or more machine-readable storage media store machine executable instructions that can be executed by the one or more processors; and

the one or more processors are configured to execute the machine executable instructions to implement the method according to claim 12 .

15 . A non-transitory machine readable storage medium, wherein the non-transitory machine readable storage medium stores computer instructions that are executable by one or more processors to implement the method according to claim 1 .

16 . A non-transitory machine readable storage medium, wherein the machine readable storage medium stores computer instructions, and one or more processors execute the computer instructions to implement the method according to claim 12 .

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (24)