SYSTEM AND METHOD OF DETERMINING OBJECT POSE

• US 20080069435A1
• Filed: 09/19/2007
• Published: 03/20/2008
• Est. Priority Date: 09/19/2006
• Status: Active Grant

• Current Assignee: Roboticvisiontech LLC
• Original Assignee: Roboticvisiontech LLC
• Inventors: Beis, Jeffrey, Boca, Remus, Pescaru, Simona, Habibi, Babak
• Granted Patent: US 8,437,535 B2
• Time in Patent Office: Days
• US Class Current: 382/153
• CPC Class Codes: G01B 11/25 by projecting a pattern, e.g. one or more lines, moiré fringes on the object G01B11/255 takes precedence ; image analysis for depth or shape recovery G06T7/50
• CPC Class Codes: G01S 17/875 for determining attitude

Abstract

Briefly described, one embodiment determines pose of an object of interest at a run time by capturing a first image of a first structured light pattern projected onto a first local surface of the object of interest; determining a first run-time data set from the captured first image, wherein the first run-time data set corresponds to information determined from the first structured light pattern projected onto the first local surface; comparing the determined first run-time data set and a corresponding first reference data set, the first reference data set corresponding to an ideal pose of the first local surface on an ideally posed reference object; and determining at least one first degree of constraint that defines a first partial pose of the first local surface, the at least one first degree of constraint based upon the comparison of the first run-time data set with the corresponding first reference data set.

Claims

1. A method for determining a pose of an object of interest at a run time, the method comprising:

capturing a first image of a first structured light pattern projected onto a first local surface of the object of interest;

determining a first run-time data set from the captured first image, wherein the first run-time data set corresponds to information determined from the first structured light pattern projected onto the first local surface;

comparing the determined first run-time data set and a corresponding first reference data set, the first reference data set corresponding to an ideal pose of the first local surface on an ideally posed reference object; and

determining at least one first degree of constraint that defines a first partial pose of the first local surface, the at least one first degree of constraint based upon the comparison of the first run-time data set with the corresponding first reference data set.

2. The method of claim 1 wherein the first local surface is a portion of a surface on the object of interest, and wherein the first local surface is characterized by absence of a discernable feature that, when captured in the first image, would otherwise provide sufficient information to determine a complete pose of the object of interest.

3. The method of claim 1, further comprising:

capturing a second image of a second structured light pattern projected onto a second local surface of the object of interest;

determining a second run-time data set from the captured second image, wherein the second run-time data set corresponds to information determined from the second structured light pattern projected onto the second local surface;

comparing the determined second run-time data set and a corresponding second reference data set, the second reference data set corresponding to the ideal pose of the second local surface on the ideally posed reference object; and

determining at least one second degree of constraint that defines a second partial pose of the second local surface, the at least one second degree of constraint based upon the comparison of the second run-time data set and the corresponding second reference data set.

4. The method of claim 3, further comprising:

determining the pose of the object of interest based upon the at least one first degree of constraint that defines the first partial pose of the first local surface and the at least one second degree of constraint that defines the second partial pose of the second local surface.

5. The method of claim 3, further comprising:

combining the at least one first degree of constraint that defines the first partial pose of the first local surface with the at least one second degree of constraint that defines the second partial pose of the second local surface; and

determining the pose of the object of interest based upon the combined at least one first degree of constraint and the at least one second degree of constraint.

6. The method of claim 3 wherein comparing the determined first run-time data set and the corresponding first reference data set, determining the at least one first degree of constraint, comparing the determined second run-time data set and the corresponding second reference data set, and determining the at least one second degree of constraint is replaced by the method comprising:

combining the first run-time data set with the second run-time data set to determine a global data set;

comparing the global data set and a corresponding reference global data set, the reference global data set data set corresponding to the ideal pose of at least the first local surface and the second local surface on the ideally posed reference object; and

determining a plurality of degrees of constraint that define at least partial pose of the first local surface and the second local surface.

7. The method of claim 6, further comprising:

determining the pose of the object of interest based upon the determined plurality of degrees of constraint.

8. The method of claim 3, further comprising:

moving an image capture device after capturing the first image so that the second image is captured with a different orientation.

9. The method of claim 1, further comprising:

determining a transformation that defines a spatial relationship between the first run-time data set and the corresponding first reference data set.

10. The method of claim 1, further comprising:

projecting the first structured light pattern onto the first local surface of the object of interest.

11. The method of claim 10 wherein projecting the first structured light pattern comprises:

projecting a first structured laser light pattern.

12. The method of claim 10 wherein projecting the first structured light pattern comprises:

projecting a first moire fringe pattern.

13. The method of claim 1, further comprising:

capturing a second image of the first structured light pattern projected onto the first local surface of the object of interest; and

determining a second run-time data set from the captured second image, wherein the second run-time data set corresponds to information determined from the first structured light pattern projected onto the first local surface.

14. The method of claim 13, further comprising:

comparing the determined first run-time data set and the second run-time data set to determine an aggregate run-time data set;

comparing the determined aggregate run-time data set and a corresponding reference data set, the reference data set corresponding to an ideal pose of the first local surface on an ideally posed reference object; and

determining at least one first degree of constraint that defines a first partial pose of the first local surface, the at least one first degree of constraint based upon the comparison of the aggregate run-time data set with the corresponding first reference data set.

15. The method of claim 13, further comprising:

comparing the determined second run-time data set and a corresponding reference data set, the reference data set corresponding to an ideal pose of the first local surface on an ideally posed reference object;

determining at least one second degree of constraint that defines the first partial pose of the first local surface, the at least one second degree of constraint based upon the comparison of the second run-time data set with the corresponding first reference data set;

comparing the determined first degree of constraint and the second degree of constraint to determine an aggregate degree of constraint; and

determining at least partial pose of the first local surface based upon the comparison of the a first degree of constraint and the second degree of constraint.

16. The method of claim 1, further comprising:

determining a second degree of constraint that further defines the first partial pose of the first local surface, the at least one second degree of constraint based upon the comparison of the first run-time data set with the corresponding first reference data set.

17. The method of claim 10 wherein projecting the first structured light pattern comprises:

projecting the first structured light pattern using an optical element operable to project the first structured light pattern at an orientation of interest onto the first local surface.

18. A system that determines a pose of an object of interest at a run time, comprising:

an image capture system operable to capture a first image of a first structured light pattern projected onto a first local surface of the object of interest and operable to capture a second image of a second structured light pattern projected onto a second local surface of the object of interest;

a structured light source system operable to project the first structured light pattern onto the first local surface and operable to project the second structured light pattern onto the second local surface; and

a processor communicatively coupled to the image capture system and the structured light source system, and operable to;

determine a first run-time data set from the captured first image, wherein the first run-time data set corresponds to information determined from the first structured light pattern projected onto the first local surface;

determine a second run-time data set from the captured second image, wherein the second run-time data set corresponds to information determined from the second structured light pattern projected onto the second local surface;

compare the determined first run-time data set and a corresponding first reference data set, the first reference data set corresponding to an ideal pose of the first local surface on an ideally posed reference object, to determine at least a first degree of constraint corresponding to the pose of the first local surface; and

compare the determined second run-time data set and a corresponding second reference data set, the second reference data set corresponding to the ideal pose of the second local surface on the ideally posed reference object, to determine at least a second degree of constraint corresponding to the pose of the second local surface.

19. The system of claim 18, wherein the image capture system comprises:

a robot device operable to maneuver at least one robot device member through a workspace;

a single image capture device coupled to the at least one robot device member such that the robot device moves the single image capture device to a first position and a first orientation so that the single image capture device captures the first image, and such that the robot device moves the single image capture device to a second position and a second orientation so that the single image capture device captures the second image.

20. The system of claim 18, wherein the single image capture system comprises:

a first image capture device at a first position and a first orientation so that the first image capture device captures the first image; and

a second image capture device at a second position and a second orientation so that the second image capture device captures the second image.

21. The system of claim 18, wherein the structured light source system comprises:

a robot device operable to maneuver at least one robot device member through a workspace;

a single structured light source coupled to the at least one robot device member such that the robot device moves the structured light source system to a first position and a first orientation so that the structured light source system projects the first structured light pattern onto the first local surface, and such that the robot device moves the structured light source system to a second position and a second orientation so that the structured light source system projects the second structured light pattern onto the second local surface.

22. The system of claim 18, wherein the structured light source system comprises:

a first structured light source at a first position and a first orientation so that the first structured light source projects the first structured light pattern onto the first local surface; and

a second structured light source at a second position and a second orientation so that the second structured light source projects the second structured light pattern onto the second local surface.

23. The system of claim 18, wherein the structured light source system comprises:

a laser light source, such that the laser light source at least projects a first structured laser light pattern onto the first local surface.

24. A method for determining a pose of an object of interest, the method comprising:

capturing at least one reference image of each of a plurality of reference local surfaces on a reference object oriented in a reference pose, wherein each of the plurality of reference local surfaces of the reference object has a structured light pattern projected thereon; and

determining a plurality of reference data sets, wherein one reference data set is determined for each one of the plurality of reference local surfaces, and wherein the plurality of reference data sets corresponds to information determined from the structured light pattern projected onto its respective reference local surface.

25. The method of claim 24, further comprising:

determining at least one reference degree of constraint from each of the plurality of reference data sets; and

determining a partial reference pose for each of the reference local surfaces based upon the determined at least one reference degree of constraint.

26. The method of claim 24, further comprising:

capturing at least one image of each of a plurality of local surfaces on the object of interest in an unknown pose at a run time, wherein each of the local surfaces of the object of interest in the unknown pose has the structured light pattern projected thereon, and wherein each one of the local surfaces of the object of interest in the unknown pose correspond to one of the reference local surfaces;

determining a plurality of run-time data sets, wherein one run-time data set is determined for each one of the plurality of local surfaces, and wherein the run-time data sets correspond to information determined from the structured light pattern projected onto its respective local surface; and

comparing the determined run-time data sets and the corresponding reference data sets.

27. The method of claim 26, further comprising:

determining at least one reference degree of constraint from each of the reference data sets; and

determining at least one run-time degree of constraint from each of the run-time data sets,

wherein comparing the run-time data sets and the reference data sets comprises;

comparing the determined at least one run-time degree of constraint and the determined respective reference degree of constraint; and

determining the pose of the object of interest in the unknown pose based upon the comparison of the determined at least one run-time degree of constraint and the determined respective reference degree of constraint.

28. The method of claim 27, further comprising:

determining a reference partial pose for each of the reference local surfaces based upon the determined at least one reference degree of constraint; and

determining a run-time partial pose for each of the local surfaces based upon the determined at least one run-time degree of constraint,

wherein comparing the run-time data sets with the reference data sets comprises;

comparing the determined run-time partial pose and the determined respective reference partial pose; and

determining the pose of the object of interest in the unknown pose based upon the comparison of the determined run-time partial pose with the determined respective reference partial pose.

29. A system that determines a pose of an object of interest at a run time, comprising:

means for capturing at least one image of each of a plurality of local surfaces on the object of interest in an unknown pose at the run time, wherein each of the local surfaces of the object of interest in the unknown pose has a structured light pattern projected thereon, and wherein each one of the local surfaces of the object of interest in the unknown pose corresponds to one of a plurality of reference local surfaces;

means for determining a plurality of run-time data sets, wherein one run-time data set is determined for each one of the plurality of local surfaces, and wherein the plurality of run-time data sets corresponds to information determined from the structured light pattern projected onto its respective local surface; and

means for comparing the determined run-time data sets and a plurality of corresponding reference data sets.

30. The system of claim 29, further comprising:

means for capturing at least one reference image of each of a plurality of reference local surfaces on a reference object oriented in a reference pose, wherein each of the plurality of reference local surfaces of the reference object has a structured light pattern projected thereon; and

means for determining the plurality of reference data sets, wherein one reference data set is determined for each one of the plurality of reference local surfaces, and wherein the plurality of reference data sets corresponds to information determined from the structured light pattern projected onto its respective reference local surface.