Method and apparatus for recognizing 3-D objects

• US 20060013450A1
• Filed: 06/22/2005
• Published: 01/19/2006
• Est. Priority Date: 06/22/2004
• Status: Active Grant

• Current Assignee: SRI International, Inc.
• Original Assignee: SRI International, Inc.
• Inventors: Kumar, Rakesh, Shan, Ying, Sawhney, Harpreet Singh, Mihai Matei, Bogdan Calin
• Granted Patent: US 8,345,988 B2
• Time in Patent Office: Days
• US Class Current: 382/123
• CPC Class Codes: G06V 10/757 Matching configurations of points or features
• CPC Class Codes: G06V 20/64 Three-dimensional objects
• CPC Class Codes: G06V 20/653 by matching three-dimensional models, e.g. conformal mapping of Riemann surfaces

Abstract

A method and apparatus for recognizing an object, comprising providing a set of scene features from a scene, pruning a set of model features, generating a set of hypotheses associated with the pruned set of model features for the set of scene features, pruning the set of hypotheses, and verifying the set of pruned hypotheses is provided.

Claims

1. A method for recognizing an object, comprising:

providing a set of scene features from a scene;

pruning a set of model features;

generating a set of hypotheses associated with said pruned set of model features for said set of scene features;

pruning said set of hypotheses; and

verifying said set of pruned hypotheses.

2. The method of claim 1, wherein at least one of said set of hypotheses is used to transform said scene features into said pruned model features.

3. The method of claim 1, wherein said step of pruning a set of model features comprises applying surface descriptor parameters to prune bogus model features.

4. The method of claim 3, wherein said applying step comprises applying locality sensitive hashing (LSH) to prune said bogus model features.

5. The method of claim 1, wherein said step of pruning said set of hypotheses comprises applying doublet-based hypotheses parameters to remove bogus hypotheses.

6. The method of claim 5, wherein said step of applying doublet-based hypotheses parameters comprises:

obtaining a pair of scene features for transforming to a corresponding pair of model features; and

generating a pose hypothesis.

7. The method of claim 6, wherein said step of applying doublet-based hypotheses parameters comprises associating angle and distance between each sets of pairs.

8. The method of claim 1, wherein said step of verifying said set of pruned hypotheses comprises determining a likelihood of at least one hypothesis.

9. The method of claim 8, wherein said step of determining said likelihood of at least one hypothesis comprises applying a RANSAC-based verification method.

10. The method of claim 9, wherein said RANSAC-based verification method is a multi-model RANSAC-based verification method.

11. A computer-readable medium having stored thereon a plurality of instructions, which, when executed by a processor, cause the processor to perform the steps of a method for recognizing 3-D objects, comprising:

providing a set of scene features from a scene;

pruning a set of model features;

generating a set of hypotheses associated with said pruned set of model features for said set of scene features;

pruning said set of hypotheses; and

verifying said set of pruned hypotheses.

12. The computer-readable medium of claim 11, wherein at least one of said set of hypotheses is used to transform said scene features into said pruned model features.

13. The computer-readable medium of claim 11, wherein said step of pruning a set of model features comprises applying surface descriptor parameters to prune bogus model features.

14. The computer-readable medium of claim 13, wherein said applying step comprises applying locality sensitive hashing (LSH) to prune said bogus model features.

15. The computer-readable medium of claim 11, wherein said step of pruning said set of hypotheses comprises applying doublet-based hypotheses parameters to remove bogus hypotheses.

16. The computer-readable medium of claim 15, wherein said step of applying doublet-based hypotheses parameters comprises:

obtaining a pair of scene features for transforming to a corresponding pair of model features; and

generating a pose hypothesis.

17. The computer-based medium of claim 16, wherein said step of applying doublet-based hypotheses parameters comprises associating angle and distance between each sets of pairs.

18. The computer-based medium of claim 11, wherein said step of verifying said set of pruned hypotheses comprises determining a likelihood of at least one hypothesis.

19. The computer readable method of claim 18, wherein said step of determining said likelihood of at least one hypothesis comprises applying a RANSAC-based verification method.

20. An apparatus for recognizing an object, comprising:

means for providing a set of scene features from a scene;

means for pruning a set of model features;

means for generating a set of hypotheses associated with said pruned set of model features for said set of scene features;

means for pruning said set of hypotheses; and

means for verifying said set of pruned hypotheses.