System and method for effectively performing a scene representation procedure

1. A system for performing a scene representation procedure, comprising:

an image manager that processes a source image to define subscenes in said source image, said image manager creating an image understanding graph based upon said subscenes, said image manager also generating a scene representation graph based upon said subscenes, said image understanding graph and said scene representation graph being different from each other and from said source image; and

a processor of an electronic device that controls said image manager to perform said scene representation procedure.

2. The system of claim 1 wherein said image understanding graph includes subscene labels, subscene location information, subscene size information, and subscene spatial connectivity information.

3. A system for performing a scene representation procedure, comprising:

an image manager that processes a source image to define subscenes in said source image, said image manager creating an image understanding graph based upon said subscenes, said image manager also generating a scene representation graph based upon said subscenes, said scene representation graph including subscene labels and subscene connectivity information without subscene location information and subscene size information; and

a processor of an electronic device that controls said image manager to perform said scene representation procedure.

4. The system of claim 3 wherein said image manager merges identical subscene labels in said scene representation graph.

5. The system of claim 1 wherein said image manager creates an image understanding description based upon said image understanding graph.

6. The system of claim 1 wherein said image manager creates a scene classification description based upon said image understanding graph.

7. The system of claim 1 wherein said image manager creates an integrated scene representation to represent a plurality of source images from a selected scene.

8. The system of claim 7 wherein said integrated scene representation includes a scene definition node that identifies said selected scene.

9. The system of claim 8 wherein said scene definition node is specified by a system user.

10. The system of claim 7 wherein said integrated scene representation includes subscene nodes from said plurality of source images.

11. The system of claim 10 wherein said integrated scene representation includes one or more first-level subscene group nodes that represent adjacent pairs of said subscene nodes.

12. The system of claim 11 wherein said integrated scene representation includes at least one second-level subscene group node that represents an adjacent pair of said first-level subscene group nodes.

13. The system of claim 1 wherein said image manager creates said scene representation graph based upon said image understanding graph.

14. The system of claim 7 wherein said image manager creates said integrated scene representation by utilizing respective ones of said scene representation graph and said image understanding graph corresponding to said plurality of source images.

15. The system of claim 7 wherein said image manager assigns an uncategorized image to an appropriate scene by matching characteristics of said uncategorized image to said integrated scene representation.

16. The system of claim 1 wherein said image manager performs a segmentation procedure upon said source image to produce a segmented image.

17. The system of claim 16 wherein said image manager assigns subscene labels to said segmented image by utilizing local human knowledge information in an image database.

18. The system of claim 17 wherein said image manager performs a subscene rectification procedure upon said subscene labels by utilizing global human knowledge information in a semantic database.

19. The system of claim 18 wherein said image manager performs a semantic refinement procedure by merging at least one similar pair of said subscene labels.

20. A method for performing a scene representation procedure, comprising the steps of:

processing a source image to define subscenes in said source image;

creating an image understanding graph based upon said subscenes and

generating a scene representation graph based upon said subscenes, said image understanding graph and said scene representation graph being different from each other and from said source image.