Method and system for performing image re-identification

1. A method for re-identifying an image, comprising:

obtaining a plurality of images using a plurality of cameras;

generating with a computer a correspondence map between the plurality of images using structural information;

defining with the computer a plurality of region signatures for one or more regions comprising the plurality of images, wherein the one or more regions are represented as a graph comprising a set of vertices, a set of spatial edges and a set of temporal edges; and

comparing with the computer the plurality of images based on the correspondence map and the plurality of region signatures to perform image re-identification.

2. The method of claim 1 , wherein the plurality of images correspond to images of individuals taken across multiple disjointed fields of view.

3. The method of claim 1 , wherein said generating a correspondence map comprises mapping the plurality of images based on structural information associated with the plurality of images.

4. The method of claim 3 , wherein said generating a correspondence map further comprises segmenting the plurality of images based on a decomposable triangulated graph technique.

5. The method of claim 4 , comprising generating a model of the plurality of images based on the decomposable triangulated graph technique.

6. The method of claim 5 , comprising fitting the generated models to the plurality of images based on a dynamic programming technique.

7. The method of claim 1 , wherein said defining a plurality of region signatures comprises generating a feature vector for the one or more regions comprising the plurality of images.

8. The method of claim 7 , wherein the feature vector comprises a histogram of a hue value and a saturation value associated with the one or more regions and a histogram of a plurality of edgels associated with the one or more regions.

9. The method of claim 1 , wherein said defining a plurality of region signatures further comprises performing a spatiotemporal segmentation of the one or more regions comprising the plurality of images.

10. The method of claim 9 , wherein said performing a spatiotemporal segmentation of the one or more regions comprises defining the plurality of region signatures based on a plurality of structural information associated with the one or more regions.

11. The method of claim 10 , wherein said performing a spatiotemporal segmentation comprises generating a plurality of spatiotemporally similar regions corresponding to the plurality of images based on a graph partitioning algorithm and comparing the plurality of images based on the plurality of spatiotemporally similar regions.

12. A method for re-identifying an image, comprising:

obtaining a plurality of images using a plurality of cameras;

defining with a computer a plurality of region signatures for one or more regions comprising the plurality of images, wherein the one or more regions are represented as a graph comprising a set of vertices, a set of spatial edges and a set of temporal edges;

performing with the computer a spatiotemporal segmentation of the one or more regions comprising the plurality of images;

generating with the computer a plurality of spatiotemporally similar regions corresponding to the plurality of images; and

comparing with the computer the plurality of images, based on the plurality of spatiotemporally similar regions, to perform image re-identification.

13. The method of claim 12 , wherein said defining a plurality of region signatures comprises generating a feature vector for the one or more regions comprising the plurality of images.

14. The method of claim 13 , wherein the feature vector comprises a histogram of a hue value and a saturation value associated with the one or more regions and a histogram of a plurality of edgels associated with the one or more regions.

15. The method of claim 12 , wherein said performing a spatiotemporal segmentation of the one or more regions comprises defining the plurality of region signatures based on a plurality of structural information associated with the one or more regions.

16. The method of claim 15 , wherein said performing a spatiotemporal segmentation of the one or more regions comprises performing an oversegmentation of one or more foreground regions comprising the plurality of images.

17. The method of claim 12 , wherein the one or more regions are represented as a graph comprising a set of vertices and a set of edges.

18. The method of claim 17 , wherein the set of edges comprise a set of spatial edges and a set of temporal edges.

19. The method of claim 12 , wherein said generating a plurality of spatiotemporally similar regions corresponding to the plurality of images comprises using a graph partitioning algorithm.

20. The method of claim 12 , wherein said comparing the plurality of images further comprises generating a correspondence map between the plurality of images based on structural information associated with the plurality of images.

21. The method of claim 20 , wherein said generating a correspondence map comprises segmenting the plurality of images to be compared using a decomposable triangulated graph technique.

22. The method of claim 21 , comprising generating a model of the plurality of images, based on the decomposable triangulated graph technique.

23. The method of claim 22 , comprising fitting the generated models, to the plurality of images, based on a dynamic programming technique.

24. A system for re-identifying an image, comprising:

a plurality of cameras configured to obtain a plurality of images; and

a computer configured to process the plurality of images, wherein the computer:

generates a correspondence map between the plurality of images using structural information;

defines a plurality of region signatures for one or more regions comprising the plurality of images, wherein the one or more regions are represented as a graph comprising a set of vertices, a set of spatial edges and a set of temporal edges; and

compares the plurality of images based on the correspondence map and the plurality of region signatures to perform image re-identification.

25. The system of claim 24 , wherein the plurality of images correspond to images of individuals taken across multiple disjoint fields of view.

26. The system of claim 24 , wherein said generating a correspondence map comprises mapping the plurality of images based on structural information associated with the plurality of images.

27. The system of claim 26 , wherein the computer is configured to:

segment the plurality of images based on a decomposable triangulated graph technique;

generate a model of the plurality of images to be compared, based on the decomposable triangulated graph technique; and

fit the generated models, to the plurality of images based on a dynamic programming technique.

28. The system of claim 24 , wherein the computer is configured to define the plurality of region signatures based on a feature vector associated with the one or more regions comprising the plurality of images.

29. The system of claim 28 , wherein the feature vector comprises a histogram of a hue value and a saturation value associated with the one or more regions and a histogram of a plurality of edgels associated with the one or more regions.

30. The system of claim 24 , wherein the computer is configured to define the plurality of region signatures based on a spatiotemporal segmentation of the one or more regions comprising the plurality of images.

31. The system of claim 30 , wherein the computer is configured to perform the spatiotemporal segmentation of the one or more regions by defining the plurality of region signatures based on a plurality of structural edgel information associated with the one or more regions.

32. The system of claim 31 , wherein the computer is configured to generate a plurality of spatiotemporally similar regions corresponding to the plurality of images based on a graph partitioning algorithm and compare the plurality of images based on the plurality of spatiotemporally similar regions.