Method and system for producing relevance sorted video summary

1. A method comprising:

obtaining a source video containing a plurality of source objects;

receiving features descriptive of at least some of the source objects;

clustering the source objects into clusters, each cluster including source objects that are similar in respect to one of the features or a combination of the features;

obtaining relevance level of the clustered source objects, respectively;

generating synopsis objects by sampling respective clustered source objects; and

generating a synopsis video having an overall play time shorter than the overall play time of the source video, by determining a play time for each of the synopsis objects, based at least partially on the respective relevance level,

wherein two or more synopsis objects which are played at least partially simultaneously in the synopsis video, are generated from source objects that are captured at different times in the source video, and

wherein said relevance level indicates a distance, on a feature space, between a synopsis object and said received features descriptive of at least some of the source objects.

2. The method according to claim 1 , wherein the determining of the play time for each of the synopsis objects, based on the respective relevance level is carried out such that synopsis objects having a higher relevance level are presented earlier than synopsis objects having a lower relevance level.

3. The method according to claim 1 , wherein the determining of the play time for each of the synopsis objects, based on the respective relevance level is carried out such that clusters having a higher relevance level are presented earlier than clusters having a lower relevance level.

4. The method according to claim 1 , wherein the features descriptive of at least some of the source objects comprise: at least one of: appearance; and activity of the objects.

5. The method according to claim 4 , wherein said appearance comprises at least one of: colors of the objects and size of the objects, and wherein said activity comprises at least one of: a velocity of the objects, speed and direction of the objects, and a path of the objects.

6. The method according to claim 1 , wherein the relevance level of the clustered source objects is determined relating to a query.

7. The method according to claim 6 , wherein the query is provided by a user.

8. The method according to claim 6 , wherein the relevance level of the clustered source objects is determined by similar means used in generating the clusters.

9. The method according to claim 1 , wherein at least one cluster is determined by a similarity of source objects to a set of features selected by a user.

10. The method according to claim 9 , wherein a set of selected features is determined by the user by pointing to a sample of one or more source objects having desired features.

11. The method according to claim 1 , wherein said determining of a play time for each of the synopsis objects is further based on relevance cost which favors a display of more relevant objects before less relevant objects.

12. A system comprising:

a video memory configured to obtain a source video possibly from a video capturing device containing a plurality of source objects;

a preprocessor configured to extract source objects from the source video and store them on object memory;

a source object selector configured to obtain features descriptive of at least some of the source objects;

a clustering unit configured to cluster the source objects into clusters, each cluster including source objects that are similar in respect to one of the features or a combination of the features and to obtain relevance level of the clustered source objects, respectively;

a synopsis object sampler configured to generate synopsis objects by sampling respective clustered source objects; and

a frame generator configured to generate a synopsis video having an overall play time shorter than the overall play time of the source video, by determining a play time for each of the synopsis objects, based at least partially on the respective relevance level, wherein two or more synopsis objects which are played at least partially simultaneously in the synopsis video, are generated from source objects that are captured at different times in the source video, wherein said relevance level indicates a distance, on a feature space, between a synopsis object and said received features descriptive of at least some of the source objects.

13. The system according to claim 12 , wherein the determining of the play time for each of the synopsis objects, based on the respective relevance level is carried out such that synopsis objects having a higher relevance level are presented earlier than synopsis objects having a lower relevance level.

14. The system according to claim 12 , wherein the determining of the play time for each of the synopsis objects, based on the respective relevance level is carried out such that clusters having a higher relevance level are presented earlier than clusters having a lower relevance level.

15. The system according to claim 12 , wherein the features descriptive of at least some of the source objects comprise: at least one of: appearance; and activity of the objects.

16. The system according to claim 12 , wherein the relevance level of the clustered source objects is determined relating to a query.

17. The system according to claim 16 , wherein the query is provided by a user.

18. The system according to claim 16 , wherein the relevance level of the clustered source objects is determined by similar means used in generating the clusters.