Method and apparatus for testing stereo vision methods using stereo imagery data

• US 20050275717A1
• Filed: 06/10/2005
• Published: 12/15/2005
• Est. Priority Date: 06/10/2004
• Status: Abandoned Application

• Current Assignee: SRI International, Inc.
• Original Assignee: Sarnoff Corporation (SRI International, Inc.)
• Inventors: Camus, Theodore Armand
• Application Number:
  US11/149,687
• Publication Number:
  US 20050275717A1
• Time in Patent Office: Days
• US Class Current: 348/42
• CPC Class Codes: G06T 2207/10012 Stereo images
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: H04N 13/275 from 3D object models, e.g. computer-generated stereoscopic image signals
• CPC Class Codes: H04N 17/00 Diagnosis, testing or measuring for television systems or their details

Abstract

A method and apparatus for generating stereo imagery scenario data to be used to test stereo vision methods such as detection, tracking, classification, steering, collision detection and avoidance methods is provided.

Claims

1. A method for testing a stereo vision method, comprising:

generating stereo imagery scenario data; and

linking the generated stereo imagery scenario data to the stereo vision method under test.

2. The method of claim 1, wherein the stereo vision method comprises at least one of:

a detection method, a tracking method, a classification method, a steering method, a collision detection method, or an avoidance method.

3. The method of claim 2, wherein the generating step comprises populating at least one scene with at least one of:

a computer generated relational object, or a computer generated relational parameter.

4. The method of claim 3, wherein the computer generated relational objects are selected from a group comprising at least one of:

a vehicle model, a road sign, a barrier, a street, an intersection, a building, a tree, a utility pole, a pedestrian, a bicyclist, or a background object.

5. The method of claim 3, wherein the computer generated relational parameter is selected from a group comprising at least one of:

an object position, an object direction, an object trajectory, an object texture, a scene illumination or a scene environmental condition.

6. The method of claim 1, wherein the generating step comprises:

generating stereo vision 3-D images; and

generating simulated scenario data from the generated stereo vision 3-D images using an image stream sequencer.

7. The method of claim 6, further comprising generating a plurality of stereo imagery scenes, wherein the plurality of stereo imagery scenes is used in the step of generating stereo vision 3-D images.

8. The method of claim 7, wherein the step of generating plurality of stereo imagery scenes comprises:

importing computer generated relational objects into each scene;

mapping the relational objects with texture;

arranging the relational objects in each scene in accordance with predetermined parameters; and

relating the objects in each scene in accordance with the predetermined parameters.

9. The method of claim 6, further comprising:

validating the generated simulated scenario data.

10. The method of claim 9, wherein the validating step comprises:

generating real images under scene conditions;

generating real scenario data using a real image stream;

linking the real scenario data to the method under test; and

comparing the test results generated from use of the simulated scenario data with the test results generated from use of the real scenario data.

11. The method of claim 10, wherein the comparison step comprising using statistical image measures to validate the simulated scenario data.

12. An apparatus for testing a stereo vision method, comprising:

means for generating stereo imagery scenario data; and

means for linking the generated stereo imagery scenario data to the stereo vision method under test.

13. The apparatus of claim 12, wherein the stereo vision method comprises at least one of:

a detection method, a tracking method, a classification method, a steering method, a collision detection method, or an avoidance method.

14. The apparatus of claim 13, wherein the means for generating stereo imagery scenario data comprises means for populating at least one scene with at least one of:

a computer generated relational object, or a computer generated relational parameter.

15. The apparatus of claim 14, wherein the computer generated relational object is selected from a group comprising at least on of:

a vehicle model, a road sign, a barrier, a street, an intersection, a building, a tree, a utility pole, a pedestrian, a bicyclist, or a background object.

16. The apparatus of claim 14, wherein the computer generated relational parameter is selected from a group comprising at least one of:

an object position, an object direction, an object trajectory, an object texture, a scene illumination, or a scene environmental condition.

17. The apparatus of claim 12, wherein the means for generating stereo imagery scenario data comprises:

means for generating stereo vision 3-D images; and

means for generating scenarios from the generated stereo vision 3-D images.

18. 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 testing a stereo vision method, comprising:

generating stereo imagery scenario data; and

linking the generated stereo imagery scenario data to the stereo vision method under test.

19. The computer-readable medium of claim 18, wherein the stereo vision method comprises at least one of:

a detection method, a tracking method, a classification method, a steering method, a collision detection method, or an avoidance method.

20. The computer-readable medium of claim 18, wherein the means for generating stereo imagery scenario data comprises means for populating at least one scene with at least one of:

computer generated relational objects, or computer generated relational parameters.

21. The computer-readable medium of claim 20, wherein the computer generated relational objects are selected from a group comprising at least one of:

a vehicle model, a road sign, an intersection, a barrier, a street, a building, a tree, an utility pole, a pedestrian, a bicyclist, or a background object.

22. The computer-readable medium of claim 20, wherein the computer generated relational parameter is selected from a group comprising at least one of:

an object position, an object direction, an object trajectory, an object texture, a scene illumination or a scene environmental condition.

23. The computer readable medium of claim 18, wherein the generating step comprises:

generating stereo vision 3-D images; and

generating scenarios from the generated stereo vision 3-D images using an image stream sequencer.