Systems and methods for robust video signature with area augmented matching

• US 8,380,045 B2
• Filed: 10/09/2008
• Issued: 02/19/2013
• Est. Priority Date: 10/09/2007
• Status: Active Grant

• Current Assignee: TiVo Solutions Inc. (Adeia Inc.)
• Original Assignee: Matthew G. Berry, Schuyler E. Eckstrom
• Inventors: Berry, Matthew G., Eckstrom, Schuyler E.
• Primary Examiner(s): Tran, Thai
• Assistant Examiner(s): HASAN, SYED Y
• Application Number:
  US12/248,596
• Publication Number:
  US 20090092375A1
• Time in Patent Office: 1,594 Days
• Field of Search: 386/248, 386/257, 386/329, 386/E5.003, 348/207.1, 348/220.1, 348/222.1, 348/231.1, 348/E17.001, 348/E17.003, 348/E7.056, 375/240.01, 375/240.12, 375/240.27, 375/E7.005, 375/E7.012, 375/E7.134, 375/E7.138, 375/E7.162, 375/E7.163, 375/E7.179, 375/E7.181, 375/E7.183, 375/E7.211, 382/236, 382/278, 709/231, 715/234, 725/112, 725/113, 725/135, 725/31, 706/45
• US Class Current: 386/248
• CPC Class Codes: H04N 2005/91385 the video frame number being modulated
• CPC Class Codes: H04N 5/147 Scene change detection
• CPC Class Codes: H04N 5/913 for scrambling ; for copy protection

Abstract

Systems and methods are provided for generating unique signatures for digital video files to locate video sequences within a video file comprising calculating a frame signature for each frame of a first video; and for a second video: calculating a frame signature for each frame of the second video for corresponding first video frame signatures, calculating a frame distance between each of the corresponding video frame signatures, determining video signature similarity between the videos, and searching within a video signature similarity curve to determine a maximum corresponding to the first video within the second video. The method further applies area augmentation to the video signature similarity curve to determine a maximum from among a plurality of maxima corresponding to the first video file within the second video file.

Claims

1. A method for generating unique signatures for digital video files for use in locating video sequences within a digital video file, the method comprising the steps of:

receiving video files at a video preprocessor;

calculating a first video frame signature in the video preprocessor for each frame of a first video file;

for a second video file;

(a) setting an initial start frame to correspond to a first frame of the second video file;

(b) from the initial start frame, calculating a second video frame signature for each frame of the second video file for which a corresponding first video frame signature is present;

(c) calculating a frame distance between each second video frame signature and each corresponding first video frame signature;

(d) determining video signature similarity between the second video file and the first video file;

(e) incrementing the initial start frame to a next successive frame within the second video file; and

(f) if total frame count of the first video file does not exceed a frame count from the initial start frame to a last frame of the second video file, continuing with step (b); and

searching within a video signature similarity curve that includes each video signature similarity determination to determine one or more maximum video signature peak similarities;

computing the area from the next least local minimum to the next greatest local minimum under each maximum video signature peak similarity as the video signature similarity curve area for each maximum video signature peak similarity; and

selecting the maximum video signature peak similarity associated with the largest video signature similarity curve area as corresponding to the first video file within the second video file.

2. The method of claim 1, wherein the video signature similarity corresponds to the initial start frame and is a percentage of second frame signatures within a predetermined threshold of corresponding first frame signatures.

3. The method of claim 1, further comprising re-sampling the first video and the second video to establish a constant frame rate.

4. The method of claim 3, wherein the constant frame rate is 5 frames per second.

5. A method for generating unique signatures for digital video files for use in locating video sequences within a digital video file, the method comprising the steps of:

segmenting a first video file into a plurality of video segments in a video preprocessor;

for each video segment;

calculating a segment video frame signature for each frame of the video segment;

for a second video file;

(a) setting an initial start frame to correspond to a first frame of the second video file;

(b) from the initial start frame, calculating a second video frame signature for each frame of the second video file for which a corresponding segment video frame signature is present;

(c) calculating a frame distance between each second video frame signature and each corresponding segment video frame signature;

(d) determining video signature similarity between the second video file and the video segment;

(e) incrementing the initial start frame to a next successive frame within the second video file; and

(f) if total frame count of the video segment does not exceed a frame count from the initial start frame to a last frame of the second video file, continuing with step (b); and

searching within a video signature similarity curve that includes each video signature similarity determination to determine one or more maximum video signature peak similarities; and

computing the area from the next least local minimum to the next greatest local minimum under each maximum video signature peak similarity as the video signature similarity curve area for each maximum video signature peak similarity; and

selecting the maximum video signature peak similarity associated with the largest video signature similarity curve area as identifying a located segment from within the first video file that appears within the second video file.

6. The method of claim 5, wherein the video signature similarity corresponds to the initial start frame and is a percentage of segment frame signatures within a predetermined threshold of corresponding first frame signatures.

7. The method of claim 5, further comprising re-sampling the first video and the second video to establish a constant frame rate.

8. The method of claim 7, wherein the constant frame rate is 5 frames per second.

9. A method for generating unique signatures for digital video files for use in locating video sequences within a digital video file, the method comprising the steps of:

(i) selecting a video segment from a first video file within a video preprocessor, wherein the video segment comprises;

an initial segment frame corresponding to a first frame of the first video file; and

a predetermined number of frames;

(ii) calculating a segment video frame signature for each frame of the video segment;

(iii) for a second video file;

(a) setting an initial start frame to correspond to a first frame of the second video file;

(b) from the initial start frame, calculating a second video frame signature for each frame of the second video file for which a corresponding segment video frame signature is present;

(c) calculating a frame distance between each second video frame signature and each corresponding segment video frame signature;

(d) determining video signature similarity between the second video file and the video segment;

(e) incrementing the initial start frame to a next successive frame within the second video file; and

(f)if total frame count of the video segment does not exceed a frame count from the initial start frame to a last frame of the second video file, continuing with step (b);

(iv) searching within a video signature similarity curve that includes each video signature similarity determination to determine one or more maximum video signature peak similarities; and

computing the area from the next least local minimum to the next greatest local minimum under each maximum video signature peak similarity as the video signature similarity curve area for each maximum video signature peak similarity; and

selecting the maximum video signature peak similarity associated with the largest video signature similarity curve area as corresponding to the first video file within the second video file;

(v) incrementing the initial segment frame to correspond to a next successive frame within the first video file; and

(vi) if the predetermined number of frames does not exceed a frame count from the initial segment frame to a last frame of the first video file, then continuing with step (ii),

wherein the maximum similarity and associated maximum curve area identifies a located segment from within the first video file that appears within the second video file.

10. The method of claim 9, wherein the video signature similarity corresponds to the initial start frame and is a percentage of second frame signatures within a predetermined threshold of corresponding first frame signatures.

11. The method of claim 9, further comprising re-sampling the first video and the second video to establish a constant frame rate.

12. The method of claim 11, wherein the constant frame rate is 5 frames per second.

13. The method of claim 9, wherein the video segment has a length of 10 seconds.

14. The method of claim 9, wherein the video segment has a length between 10 seconds and 30 seconds.