Scene change detection by segmentation analysis

• US 20030228056A1
• Filed: 06/10/2002
• Published: 12/11/2003
• Est. Priority Date: 06/10/2002
• Status: Active Grant

• Current Assignee: Altera Corporation (Intel Corporation)
• Original Assignee: Pulsent Corp.
• Inventors: Prakash, Adityo, Fodor, Eniko, Ratner, Edward R.
• Granted Patent: US 6,985,623 B2
• Time in Patent Office: Days
• US Class Current: 382/173
• CPC Class Codes: G11B 27/034 on discs G11B27/036, G11B27/038 take precedence
• 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: H04N 5/147 Scene change detection

Abstract

A process and apparatus for identifying abrupt cuts or scene changes in any ordered sequence of images. In one specific embodiment, two or more consecutive images from a sequence are introduced to a segmenter as digital frames. The segmenter independently divides each of these frames into pixel regions or segments according to some common characteristic so that every pixel belongs to exactly one segment. A segment analysis unit then performs some statistical analysis on the segment data for each of the frames and generates composite statistics for each frame. A frame comparison unit then examines these composite statistics to determine whether these frames belong to a consistent scene of images. If the composite statistics for these frames differ sufficiently, the comparison unit declares the latter frame in the sequence to belong to a new scene. This information may then be transmitted back to the data source for the purpose of marking the scene change or for any other purpose.

Claims

1. A method for detecting large-scale visual discontinuities in a temporal sequence of images, the method comprising:

dividing each image of the sequence into non-overlapping segments;

compiling statistical information about the segments for each image in the sequence;

comparing the statistical information for consecutive images in the temporal sequence using a statistical measurement; and

identifying points in the temporal sequence at which abrupt changes in the statistical measurement comprise scene changes.

2. The method of claim 1 wherein the statistical information about the segments for each image in the sequence comprises a histogram.

3. The method of claim 2 wherein the histogram records a number of segments in each of a plurality of pixel size ranges.

4. The method of claim 2 wherein the histogram records a number of segments in each of a plurality of color ranges.

5. The method of claim 2 wherein the histogram records a number of segments in each of a plurality of bins where each bin includes a range of pixel sizes and a range of color frequencies.

6. The method of claim 1 wherein the comparison of statistical information for consecutive images comprises computing a chi-squared statistic.

7. The method of claim 1 wherein the comparison of statistical information for consecutive images comprises computing an nth statistical moment for each image.

8. The method of claim 1 wherein the method is performed by a video encoder for a video communication system.

9. An apparatus for detecting scene changes in a sequence of video frames, the apparatus comprising:

a segmenter for dividing the video frames into non-overlapping segments;

a segmentation analysis unit for compiling statistical information about the segments in the video frames; and

a frame comparison unit for comparing the statistical information for temporally consecutive video frames using a statistical measurement and for identifying frames at which an abrupt change in the statistical measurement is indicative of a scene change.

10. The apparatus of claim 9 wherein the statistical information about the segments in a frame comprises a histogram.

11. The apparatus of claim 10 wherein the segmentation analysis unit records a number of segments in each of a plurality of pixel size ranges in generating the histogram.

12. The apparatus of claim 11 wherein the frame comparison unit compares a chi-squared statistic of a first histogram of a frame with a chi-squared statistic of a second histogram of a next frame.

13. The apparatus of claim 10 wherein the segmentation analysis unit records a number of segments in each of a plurality of color ranges in generating the histogram.

14. The apparatus of claim 10 wherein the segmentation analysis unit records a number of segments in each of a plurality of bins, where each bin includes a range of pixel sizes and a range of color frequencies, in generating the histogram.

15. The apparatus of claim 10 wherein the statistical information comprises an n^th statistical moment for the histogram.

16. The apparatus of claim 9 wherein the apparatus comprises a video encoder for a video communication system.

17. An apparatus for detecting scene changes in a sequence of video frames, the apparatus comprising:

means for dividing the video frames into non-overlapping segments;

means for compiling statistical information about the segments in the video frames;

means for comparing the statistical information for temporally consecutive video frames using a statistical measurement; and

means for identifying frames at which an abrupt change in the statistical measurement is indicative of a scene change.

18. The apparatus of claim 17 wherein the statistical information about the segments in a frame comprises a histogram.

19. The apparatus of claim 18 wherein the histogram records a number of segments in each of a plurality of pixel size ranges.

20. The apparatus of claim 18 wherein the histogram records a number of segments in each of a plurality of color ranges.