Detecting segments of a video program

Patent metadata

Field	Value
Publication number	US-9418296-B1
Application number	US-201514660894-A
Country	US
Kind code	B1
Filing date	Mar 17, 2015
Priority date	Mar 17, 2015
Publication date	Aug 16, 2016
Grant date	Aug 16, 2016

Abstract

In an embodiment, a data store storing a first video and a second video that is associated with the first video; a computer processor coupled to the data store and programmed to: generate a first model fingerprint of the first video, based on pixels in a first model frame in a first model segment of the first video stored in the data store; generate a first test fingerprint of the second video based on pixels in a first test frame in the second video stored in the data store; determine a first closeness value between the first model fingerprint and the first test fingerprint; determine, based on the first closeness value, whether the first test frame is a first boundary of a first segment in the second video, wherein the first segment in the second video is similar to the first model segment in the first video.

First claim

What is claimed is:

1. A video processing system comprising:
a data store storing a first video and a second video that is associated with the first video;
a computer processor coupled to the data store and programmed to:
generate a first model fingerprint of the first video, based on pixels in a first model frame in a first model segment of the first video stored in the data store;
generate a first test fingerprint of the second video based on pixels in a first test frame in the second video stored in the data store;
determine a first closeness value between the first model fingerprint and the first test fingerprint;
determine, based on the first closeness value, whether the first test frame is a first boundary of a first segment in the second video, wherein the first segment in the second video is similar to the first model segment in the first video.

2. The video processing system of claim 1 , wherein the computer processor is programmed, in response to determining that the first test frame is a first boundary, to determine whether the first boundary is an ending boundary of the first segment in the second video, and if so, send a value that indicates the ending boundary to a video player on a client computer that is configured to play the second video, which value causes the video player to skip ahead to the ending boundary.

3. The video processing system of claim 1 , wherein the computer processor is programmed, in response to determining that the first test frame is a first boundary, to determine whether the first boundary is a beginning boundary of the first segment in the second video, and if so, send a value that indicates the beginning boundary to a video player on a client computer that is configured to play the second video, which value causes the video player to stop playing the second video and request a third video that is associated with the first video and begin playing the third video for a user.

4. The video processing system of claim 1 , wherein the computer processor is programmed to:
generate a second model fingerprint based on pixels in a second model frame in the first model segment of the first video stored in the data store;
generate a second test fingerprint based on pixels in a second test frame in the second video stored in the data store;
determine a second closeness value between the second model fingerprint and the second test fingerprint;
determine, based on the second closeness value, whether the second test frame is a second boundary of the first segment in the second video, wherein the first boundary is a beginning boundary of the first segment in the second video and the second boundary is an ending boundary of the first segment in the second video;
send, in response to determining that the first test frame is the first boundary and the second test frame is the second boundary, a first value that indicates the beginning boundary and a second value that indicates the ending boundary to a video player on a client computer, which causes the video player to skip ahead to the ending boundary in the second video when the video player reaches the beginning boundary.

5. The video processing system of claim 1 , wherein the computer processor is programmed to:
generate the first model fingerprint by determining a first model color distribution based on a first set of model pixels in the first model frame, wherein each particular color in the first model color distribution is associated with a value that indicates how many pixels in the first set of model pixels are assigned the particular color;
generate the first test fingerprint by determining a first test color distribution based on a first set of test pixels in the first test frame, wherein each particular color in the first test color distribution is associated with a value that indicates how pixels in the first set of test pixels are assigned the particular color.

6. The video processing system of claim 5 , wherein the computer processor is programmed to:
determine a set of difference values, wherein each difference value in the set of difference values corresponds with a color and indicates how many pixels are assigned the color in the first model color distribution compared to how many pixels are assigned the color in the first test color distribution;
determine a sum by adding each difference value in the set of difference values;
determine the first closeness value by dividing the sum by how many pixels are in the first set of model pixels.

7. The video processing system of claim 5 , wherein the computer processor is programmed to store the first model color distribution in the data store as the first model fingerprint.

8. The video processing system of claim 5 , wherein the computer processor is programmed to:
convert one or more first color components of each pixel in a first color space from the first set of model pixels and the first set of test pixels into one or more second color components defined in a second color space, wherein the first color space and the second color space are different;
determine the first model color distribution based on the one or more second color components of each pixel in the first set of model pixels;
determine the first test color distribution based on the one or more second color components of each pixel in the first set of test pixels.

9. The video processing system of claim 1 , wherein the computer processor is programmed to:
generate a second model fingerprint based on pixels in a second model frame in the first model segment of the first video stored in the data store, wherein the second model frame is different than the first model frame;
generate a second test fingerprint based on pixels in a second test frame in the second video;
determine a second closeness value between the second model fingerprint and the second test fingerprint;
determine, based on the second closeness value, whether the first test frame is a first boundary of a first segment in the second video, wherein the first segment in the second video is similar to the first model segment in the first video.

10. The video processing system of claim 9 , wherein the computer processor is programmed to determine that the first test frame is the first boundary of the first segment in the second video if the first closeness value and the second closeness value are both below a particular threshold.

11. The video processing system of claim 1 , wherein the computer processor is programmed to:
detect whether a face is in the first test frame;
in response to determining that the face is detected in first frame, withholding data from a video player on a client computer, wherein the data indicates that the video player may skip the first segment starting at the first test frame.

12. The video processing system of claim 1 , wherein the first model frame has as many pixels as the first test frame.

13. A method for requesting video from a server computer to play on a client computer comprising:
receiving input from a user selecting a first video title, wherein the first video title is associated with a second video title, and the first video title includes one or more common video segments with the second video title;
requesting, from the server computer, a set of metadata associated with the first video title indicating one or more common segments that may be skipped;
receiving the set of metadata associated with the first video title, and in response, requesting one or more first video chunks associated with the first video title without requesting one or more second video chunks that comprise frames included in the one or more common video segme

Assignees

• Netflix Inc

Inventors

• Kansara Apurvakumar

Classifications

• H04N21/23418 Primary

  involving operations for analysing video streams, e.g. detecting features or characteristics (television picture signal circuitry for scene change detection H04N5/147; filtering for image enhancement G06T5/00; methods or arrangements for recognising scenes G06V20/00; arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59) · CPC title

• G06V40/161

  Detection; Localisation; Normalisation · CPC title

• G06V10/75

  Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries · CPC title

• G06V20/46 Primary

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

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title