Vision-based method for rapid directed area search

• US 8,437,558 B1
• Filed: 10/08/2009
• Issued: 05/07/2013
• Est. Priority Date: 10/08/2009
• Status: Expired due to Fees

• Current Assignee: HRL Laboratories LLC (The Boeing Co.)
• Original Assignee: HRL Laboratories LLC (The Boeing Co.)
• Inventors: Medasani, Swarup, Owechko, Yuri
• Primary Examiner(s): DULANEY, KATHLEEN YUAN
• Application Number:
  US12/587,642
• Time in Patent Office: 1,307 Days
• Field of Search: 382/225, 382/228, 382/103, 382/104, 382/173
• US Class Current: 382/225
• CPC Class Codes: G06V 10/462 Salient features, e.g. scale invariant feature transforms [SIFT]
• CPC Class Codes: G06V 10/757 Matching configurations of points or features

Abstract

Described is a system for rapid directed area search utilizing particle swarm optimization. The system first extracts salient regions from an input image. The system then detects regions of interest from the salient regions utilizing particle swarm optimization, wherein a swarm of software agents, or particles, cooperate to locate an objective function optima, or region of interest, in an image. A set of local feature descriptors are then extracted from the image, wherein a local feature descriptor corresponds to a neighborhood surrounding a point of interest in a region of interest in the image. Additionally, the set of local feature descriptors are clustered hierarchically into a database so that a closest match between a new input image and a stored image can be determined. Finally, the matching regions of the two images are registered to align matching regions to allow detection of changes between the images.

Claims

1. A system for rapid directed area search, the system comprising one or more processors that are configured to perform operations of:

extracting at least one salient region from a new input image;

detecting at least one region of interest from a salient region in the new input image by operating a plurality of software agents as a cooperative swarm to locate an objective function optima, wherein the objective function optima is determined according to particle swarm optimization dynamics, and wherein the objective function optima corresponds to a region of interest in the new input image;

extracting a set of local feature descriptors from the new input image;

clustering and indexing the set of local feature descriptors into a database having a plurality of nodes organized as a hierarchical structure;

detecting and matching the set of extracted local feature descriptors from the new input image with a set of extracted local feature descriptors from an old image retrieved from the database using particle swarm optimization dynamics;

calculating a registration transformation that matches the new input image and the old image;

registering at least one matching region of the new input image and the old image, wherein registering of the at least one matching region aligns the images;

detecting at least one changed region between the new input image and the old image, wherein the at least one changed region is stored or presented.

2. A system for rapid directed area search as set forth in claim 1, wherein the set of local feature descriptors are clustered hierarchically using fuzzy c-means.

3. A system for rapid directed area search as set forth in claim 2, wherein the system is further configured to perform operations of partitioning the set of local feature descriptors into a plurality of clusters with each iteration, wherein the partitioning is repeated with each iteration until the plurality of clusters are homogeneous.

4. A system for rapid directed area search as set forth in claim 3, wherein the system is further configured to perform operations of forming a plurality of index tables at each node, wherein the plurality of index tables indicate whether an image stored in the database contains a feature vector contained in a cluster in addition to a location of the image in the database.

5. A system for rapid directed area search as set forth in claim 4, wherein the system is further configured to perform operations of detecting an object of interest in the at least one changed region using particle swarm optimization dynamics.

6. A system for rapid directed area search as set forth in claim 1, wherein the system is further configured to perform operations of detecting an object of interest in the at least one changed region using particle swarm optimization dynamics.

7. A system for rapid directed area search as set forth in claim 1, wherein the system further performs operations of forming a plurality of index tables at each node, wherein the plurality of index tables indicate whether an image stored in the database contains a feature vector contained in a cluster in addition to a location of the image in the database.

8. A computer-implemented method for rapid directed area search, the method comprising an act of causing a processor to perform operations of:

extracting at least one salient region from a new input image;

detecting at least one region of interest from a salient region in the new input image by operating a plurality of software agents as a cooperative swarm to locate an objective function optima, wherein the objective function optima is determined according to particle swarm optimization dynamics, and wherein the objective function optima corresponds to a region of interest in the new input image;

extracting a set of local feature descriptors from the new input image;

clustering and indexing the set of local feature descriptors into a database having a plurality of nodes organized as a hierarchical structure;

detecting and matching the set of extracted local feature descriptors from the new input image with a set of extracted local feature descriptors from an old image retrieved from the database using particle swarm optimization dynamics;

calculating a registration transformation that matches the new input image and the old image;

registering at least one matching region of the new input image and the old image, wherein registering of the at least one matching region aligns the images;

detecting at least one changed region between the new input image and the old image, wherein the at least one changed region is stored or presented.

9. A method for rapid directed area search as set forth in claim 8, wherein the set of local feature descriptors are clustered hierarchically using fuzzy c-means.

10. A method for rapid directed area search as set forth in claim 9, wherein the method further comprises an act of partitioning the set of local feature descriptors into a plurality of clusters with each iteration, wherein the partitioning is repeated with each iteration until the plurality of clusters are homogeneous.

11. A method for rapid directed area search as set forth in claim 10, wherein the method further comprises an act of forming a plurality of index tables at each node, wherein the plurality of index tables indicate whether an image stored in the database contains a feature vector contained in a cluster in addition to a location of the image in the database.

12. A method for rapid directed area search as set forth in claim 11, wherein the method further comprises an act of detecting an object of interest in the at least one changed region using particle swarm optimization dynamics.

13. A method for rapid directed area search as set forth in claim 8, wherein the method further comprises an act of forming a plurality of index tables at each node, wherein the plurality of index tables indicate whether an image stored in the database contains a feature vector contained in a cluster in addition to a location of the image in the database.

14. A method for rapid directed area search as set forth in claim 8, wherein the method further comprises an act of detecting an object of interest in the at least one changed region using particle swarm optimization dynamics.

15. A computer program product for rapid directed area search, the computer program product comprising computer-readable instruction means stored on a non-transitory computer-readable medium that are executable by a computer having a processor for causing the processor to perform operations of:

extracting at least one salient region from a new input image;

detecting at least one region of interest from a salient region in the new input image by operating a plurality of software agents as a cooperative swarm to locate an objective function optima, wherein the objective function optima is determined according to particle swarm optimization dynamics, and wherein the objective function optima corresponds to a region of interest in the new input image;

extracting a set of local feature descriptors from the new input image;

clustering and indexing the set of local feature descriptors into a database having a plurality of nodes organized as a hierarchical structure;

detecting and matching the set of extracted local feature descriptors from the new input image with a set of extracted local feature descriptors from an old image retrieved from the database using particle swarm optimization dynamics;

calculating a registration transformation that matches the new input image and the old image;

registering at least one matching region of the new input image and the old image, wherein registering of the at least one matching region aligns the images;

detecting at least one changed region between the new input image and the old image, wherein the at least one changed region is stored or presented.

16. A computer program product for rapid directed area search as set forth in claim 15, wherein the set of local feature descriptors are clustered hierarchically using fuzzy c-means.

17. A computer program product for rapid directed area search as set forth in claim 16, wherein the computer program product further comprises instruction means for partitioning the set of local feature descriptors into a plurality of clusters with each iteration, wherein the partitioning is repeated with each iteration until the plurality of clusters are homogeneous.

18. A computer program product for rapid directed area search as set forth in claim 17, wherein the computer program product further comprises instruction means for forming a plurality of index tables at each node, wherein the plurality of index tables indicate whether an image stored in the database contains a feature vector contained in a cluster in addition to a location of the image in the database.

19. A computer program product for rapid directed area search as set forth in claim 18, wherein the computer program product further comprises instruction means for detecting an object of interest in the at least one changed region using particle swarm optimization dynamics.

20. A computer program product for rapid directed area search as set forth in claim 15, wherein the computer program product further comprises instruction means for forming a plurality of index tables at each node, wherein the plurality of index tables indicate whether an image stored in the database contains a feature vector contained in a cluster in addition to a location of the image in the database.

21. A computer program product for rapid directed area search as set forth in claim 15, wherein the computer program product further comprises instruction means for detecting an object of interest in the at least one changed region using particle swarm optimization dynamics.