Method and apparatus for processing a video

Patent metadata

Field	Value
Publication number	US-10897620-B2
Application number	US-201916567976-A
Country	US
Kind code	B2
Filing date	Sep 11, 2019
Priority date	Dec 14, 2018
Publication date	Jan 19, 2021
Grant date	Jan 19, 2021

Abstract

Embodiments of the present disclosure relate to a method and apparatus for processing a video. The method can include: acquiring a to-be-processed video, and dividing the video into at least one video segment according to a scene; for a video segment of the at least one video segment, extracting a feature of the video segment; inputting the feature and a target video quality into a pre-trained encoding parameter prediction model, to obtain an encoding parameter of the video segment, the encoding parameter prediction model being used to represent a corresponding relationship between the feature of the video segment and the encoding parameter under the target video quality; and performing an encoding process on the video according to the encoding parameter corresponding to the video segment of the at least one video segment, to generate an encoded video.

First claim

What is claimed is:

1. A method for processing a video, comprising:
acquiring a to-be-processed video, and dividing the video into at least one video segment according to a scene;
for a video segment of the at least one video segment, extracting a feature of the video segment;
inputting the feature and a target video quality into a pre-trained encoding parameter prediction model, to obtain an encoding parameter of the video segment, the encoding parameter prediction model being used to represent a corresponding relationship between the feature of the video segment and the encoding parameter under the target video quality; and
performing an encoding process on the video according to the encoding parameter corresponding to the video segment of the at least one video segment, to generate an encoded video;
wherein the encoding parameter prediction model is trained and obtained by:
acquiring a sample video set, and annotating an encoding parameter and a video quality of a sample video in the sample video set;
extracting, for the sample video in the sample video set, a feature of the sample video, and determining a complexity factor of the sample video according to the feature of the sample video;
classifying the sample video in the sample video set into at least one category according to the complexity factor; and
performing, for a category of the at least one category, a regression on the encoding parameter and the video quality of the sample video belonging to the category, to obtain a relationship curve between the encoding parameter and the video quality.

2. The method according to claim 1 , further comprising:
determining a video quality of the encoded video; and
outputting, in response to the determined video quality not matching the target video quality, predetermined information.

3. The method according to claim 1 , wherein the extracting a feature of the video segment includes:
decoding the video segment, to extract a time domain feature and/or a frequency domain feature of the video segment.

4. The method according to claim 1 , wherein the annotating an encoding parameter and a video quality of a sample video in the sample video set includes:
encoding, for the sample video in the sample video set, the sample video according to a predetermined constant quality value, and annotating the encoding parameter and the video quality during the encoding.

5. The method according to claim 1 , wherein the complexity factor is related to a pixel depth of the encoded sample video.

6. An apparatus for processing a video, comprising:
at least one processor; and
a memory storing instructions, the instructions when executed by the at least one processor, causing the at least one processor to perform operations, the operations comprising:
acquiring a to-be-processed video, and divide the video into at least one video segment according to a scene;
extracting, for a video segment of the at least one video segment, a feature of the video segment;
inputting the feature and a target video quality into a pre-trained encoding parameter prediction model, to obtain an encoding parameter of the video segment, the encoding parameter prediction model being used to represent a corresponding relationship between the feature of the video segment and the encoding parameter under the target video quality; and
performing an encoding process on the video according to the encoding parameter corresponding to the video segment of the at least one video segment, to generate an encoded video;
wherein the encoding parameter prediction model is trained and obtained by:
acquiring a sample video set, and annotating an encoding parameter and a video quality of a sample video in the sample video set;
extracting, for the sample video in the sample video set, a feature of the sample video, and determining a complexity factor of the sample video according to the feature of the sample video;
classifying the sample video in the sample video set into at least one category according to the complexity factor; and
performing, for a category of the at least one category, a regression on encoding parameter and the video quality of the sample video belonging to the category, to obtain a relationship curve between the encoding parameter and the video quality.

7. The apparatus according to claim 6 , the operations further comprising:
determining a video quality of the encoded video; and
outputting, in response to the determined video quality not matching the target video quality, predetermined information.

8. The apparatus according to claim 6 , wherein the extracting a feature of the video segment includes:
decoding the video segment, to extract a time domain feature and/or a frequency domain feature of the video segment.

9. The apparatus according to claim 6 , wherein the annotating an encoding parameter and a video quality of a sample video in the sample video set includes:
encoding, for the sample video in the sample video set, the sample video according to a predetermined constant quality value, and annotating the encoding parameter and the video quality during the encoding.

10. The apparatus according to claim 6 , wherein the complexity factor is related to a pixel depth of the encoded sample video.

11. A non-transitory computer readable medium, storing a computer program, wherein the computer program, when executed by a processor, causes the processor to perform operations, the operations comprising:
acquiring a to-be-processed video, and dividing the video into at least one video segment according to a scene;
for a video segment of the at least one video segment, extracting a feature of the video segment;
inputting the feature and a target video quality into a pre-trained encoding parameter prediction model, to obtain an encoding parameter of the video segment, the encoding parameter prediction model being used to represent a corresponding relationship between the feature of the video segment and the encoding parameter under the target video quality; and
performing an encoding process on the video according to the encoding parameter corresponding to the video segment of the at least one video segment, to generate an encoded video;
wherein the encoding parameter prediction model is trained and obtained by:
acquiring a sample video set, and annotating an encoding parameter and a video quality of a sample video in the sample video set;
extracting, for the sample video in the sample video set, a feature of the sample video, and determining a complexity factor of the sample video according to the feature of the sample video;
classifying the sample video in the sample video set into at least one category according to the complexity factor; and
performing, for a category of the at least one category, a regression on encoding parameter and video quality of sample video belonging to the category, to obtain a relationship curve between the encoding parameter and the video quality.

Assignees

• Beijing Baidu Netcom Sci & Tec

Inventors

• Xing Huaifei
• Wang Jialiang
• Shen Huifeng
• Cao Feifei

Classifications

• H04N19/103 Primary

  Selection of coding mode or of prediction mode · CPC title

• G06V10/993

  Evaluation of the quality of the acquired pattern · CPC title

• H04N19/154 Primary

  Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion (use of rate-distortion criteria H04N19/147) · CPC title

• G06V20/49

  Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes · CPC title

• G06V20/46

  Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames · CPC title