Granted Patent B2

US 8,135,222 · App. 12/544,566 · Granted Mar 13, 2012

Abstract

Claims (43)

1. An image processing method comprising:

clustering images of a set of images to generate a plurality of scenes each comprising an unordered cluster of images wherein the clustering is based on pairwise comparisons between pairs of images of the set of images;

ordering images within scenes respective to video coherence to generate ordered scenes comprising ordered sequences of images; and

generating video content as images ordered in accordance with the ordered sequences of images comprising the scenes;

wherein at least the clustering and the ordering are performed by a digital processor.

2. The image processing method as set forth in claim 1 , further comprising:

displaying or storing or streaming the video content.

3. The image processing method as set forth in claim 1 , further comprising:

generating feature vectors representative of the images of the set of images; and

generating the pairwise comparisons as comparisons of feature vectors of pairs of images.

4. The image processing method as set forth in claim 3 , wherein the generating of feature vectors incorporates metadata associated with the images of the set of images into the feature vectors.

5. The image processing method as set forth in claim 1 , wherein the clustering employs soft partitioning.

6. The image processing method as set forth in claim 1 , further comprising:

assigning a scene date for each scene based on an aggregation of timestamps of images within the scene, the generating of video content comprising ordering scenes within the video content based on the scene dates.

7. The image processing method as set forth in claim 1 , wherein the ordering of images within scenes respective to video coherence comprises:

estimating a shortest path through the images within a scene respective to pairwise comparisons between pairs of images.

8. An image processing method comprising:

clustering images of a set of images to generate a plurality of scenes each comprising an unordered cluster of images;

ordering images within scenes respective to video coherence to generate ordered scenes comprising ordered sequences of images;

generating video content as images ordered in accordance with the ordered sequences of images comprising the scenes;

selecting a display time for each image; and

generating video comprising video frames based on the video content wherein each image of the video content defines one or more video frames effective to cause the image to be displayed for the determined display time when the video is played at a predetermined frame rate;

wherein at least the clustering and the ordering are performed by a digital processor.

9. The image processing method as set forth in claim 8 , wherein the ordering of images within scenes respective to video coherence comprises:

estimating a shortest path through the images within a scene respective to pairwise comparisons between pairs of images.

10. The image processing method as set forth in claim 9 , wherein the estimating a shortest path through the images within a scene comprises:

selecting a first image for the shortest path based on timestamps of the images.

11. The image processing method as set forth in claim 8 , wherein the selecting a display time for each image comprises:

selecting a display time for each image based on a constant entropy criterion for the generated video.

12. The image processing method as set forth in claim 8 , further comprising

playing the generated video at the predetermined frame rate.

13. A non-transitory storage medium storing instructions executable by a digital processor to perform an image processing method comprising clustering images of a set of images to generate a plurality of scenes and ordering images within each scene respective to a video coherence criterion, wherein the video coherence criterion comprises an estimation of a shortest path through the images within the scene wherein the estimation employs a pairwise image comparison measure.

14. The non-transitory storage medium as set forth in claim 13 , further storing instructions executable to generate video content as images ordered in accordance with the ordered sequences of images comprising the scenes.

15. The non-transitory storage medium as set forth in claim 14 , further storing instructions executable to generate video comprising frames defined by the video content.

16. The non-transitory storage medium as set forth in claim 14 , further storing instructions executable to generate video comprising frames defined by the video content wherein display times for the images within each scene are selected to conform with a constant entropy criterion.

17. The non-transitory storage medium as set forth in claim 13 , wherein the clustering is respective to pairwise image comparisons.

18. An image processing method comprising:

automatically ordering a set of images based on a video coherence criterion comprising a shortest distance estimation wherein the automatic ordering comprises clustering the set of images to generate clusters defining a plurality of scenes and automatically ordering the images within each scene based on the video coherence criterion; and

playing the images as video comprising frames defined by the ordered set of images and played at a predetermined frame rate.

19. An image processing method comprising:

automatically ordering a set of images based on a video coherence criterion;

playing the images as video comprising frames defined by the ordered set of images and played at a predetermined frame rate; and

selecting the predetermined frame rate using a constant entropy criterion.

Assignments (10)