METHOD AND APPARATUS FOR IDENTIFYING VIDEO TRANSITIONS

• US 20110064381A1
• Filed: 09/15/2009
• Published: 03/17/2011
• Est. Priority Date: 09/15/2009
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Circlaeys, Eric M., Bronder, Kjell
• Granted Patent: US 8,582,952 B2
• Time in Patent Office: Days
• US Class Current: 386/241
• CPC Class Codes: G11B 27/28 by using information signals recorded by the same method as the main recording G11B27/22 takes precedence
• CPC Class Codes: G11B 27/322 used signal is digitally coded
• CPC Class Codes: H04H 60/37 for identifying segments of broadcast information, e.g. scenes or extracting programme ID
• CPC Class Codes: H04H 60/74 using programme related information, e.g. title, composer or interpreter
• CPC Class Codes: H04N 5/147 Scene change detection

Abstract

The disclosure addresses methods and apparatus for evaluating video presentations, particularly those displaying a succession of generally static images to identify the transition from a first image to a second image. The change of images facilitates the insertion of chapter markers which may be later used to index the content of the video and to facilitate navigation to selected content in the video. In one implementation, a difference measurement will be made between selected frames of the video presentation and evaluated with a statistical measure of the magnitude of the difference measurements, such as a Root Mean Square analysis. In preferred implementations, all processing will be done in the digital domain.

Claims

1. A method of identifying chapter markers in a video presentation, comprising the acts of:

comparing video data representative of a plurality of frames of a video presentation depicting a plurality of still images to determine a plurality of pixel-based difference measurements;

evaluating the pixel-based difference measurements at least in part through use of a Root-Mean-Square analysis;

identifying periods of stability in the video presentation through use of the results of the RMS analysis, indicating presentation of still images in the video presentation; and

establishing a chapter marker in the video presentation proximate the beginning of an identified period of stability.

2. The method of claim 1, wherein the video data comprises one or more packets of digital data representative of each frame of the video presentation.

3. The method of claim 1, further comprising the act of extracting video data from a video data file, the extracted video data representative of successive frames of the video presentation.

4. The method of claim 3, wherein the successive frames of the video presentation are non-sequential, and are extracted according to at least one pre-selected criteria.

5. The method of claim 3, further comprising the scaling of the data representative of each video frame used for the comparison.

6. The method of claim 1, further comprising the act of using the established chapter marker to index a subsequent composite video comprising the video presentation containing chapter workers and a second video presentation.

7. A method of processing digital video data, comprising the acts of:

defining a plurality of parameters for processing to be performed on a digital video data set;

extracting digital video data representative of a plurality of successive video frames represented by said digital video data set, the extracting performed in reference to at least one of the defined parameters;

processing the extracted digital video data to reduce the data set representative of each of the successive video frames;

determining a pixel-based difference blend result between immediately successive extracted video frames;

performing a Root-Mean-Square analysis of each difference blend result to determine a measure of the magnitude of each difference blend result relative to a reference; and

determining transitions of content in the digital video data set comprising a change from a first generally still image to a second generally still image through reference to the Root-Mean-Square analysis of each difference blend result.

8. The method of claim 7, further comprising the act of generating chapter markers associated with the digital video data set.

9. The method of claim 8, further comprising the act of generating a video presentation comprising a combination of the digital video data stream with chapter markers and a second digital video data set.

10. The method of claim 9, wherein the chapter markers are used to index the combination video presentation.

11. The method of claim 10, wherein the combination second digital video data set is of a video presentation substantially containing moving images.

12. A machine-readable medium containing instructions that when executed by one or more processors cause the performing of operations comprising:

comparing video data representative of a plurality of frames of a video presentation depicting a plurality of still images to determine a pixel-based difference measure;

evaluating the pixel-based difference measure at least in part through use of a Root-Mean-Square analysis;

identifying a plurality of patterns in the video data indicating presentation of still images in the video presentation; and

establishing a chapter marker in the video presentation proximate the beginning of an identified pattern.

13. The machine-readable medium of claim 12, wherein the operations further comprise:

extracting frames from the video presentation to be compared; and

reducing the resolution of the extracted frames.

14. A method of processing video data, comprising the acts of:

in the digital domain, extracting a plurality of video frames from a video data set;

in the digital domain, comparing first and second video frames to one another to determine a first difference measurement;

in the digital domain, comparing the second video frame to a third video frame to determine a second difference measurement;

performing a Root-Mean-Square analysis of the first and second difference measurements; and

determining a transition from a first still image to a second still image in said video data file, the transition determined in reference to the Root-Mean-Square analysis of the first and second difference measurements.

15. The method of claim 14, wherein the act of determining a transition from a first still image to a second still image in the video data file comprises comparing a curve comprising data points representative of the Root-Mean-Square analysis of the first and second difference measurements to a reference.