Edge-based pose detection

• US 9,456,187 B1
• Filed: 06/01/2012
• Issued: 09/27/2016
• Est. Priority Date: 06/01/2012
• Status: Expired due to Fees

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Liu, Qiang, Chang, Samuel Henry, Yao, Ning
• Primary Examiner(s): Haskins, Twyler
• Assistant Examiner(s): Haliyur, Padma
• Application Number:
  US13/486,634
• Time in Patent Office: 1,579 Days
• Field of Search: 345/663
• US Class Current: 1/1
• CPC Class Codes: G06F 3/011 Arrangements for interaction with the human body, e.g. for user immersion in virtual reality blind teaching G09B21/00
• CPC Class Codes: G06F 3/017 Gesture based interaction, e.g. based on a set of recognized hand gestures interaction based on gestures traced on a digitiser G06F3/04883
• CPC Class Codes: G06T 17/00 Three dimensional [3D] modelling, e.g. data description of 3D objects
• CPC Class Codes: G06T 2207/20164 Salient point detection; Corner detection
• CPC Class Codes: G06T 7/75 involving models
• CPC Class Codes: H04W 4/02 Services making use of location information

Abstract

Techniques are described for determining the pose of a surface based on a point set representing the surface. Corner models are matched to a contour of the point set using a difference minimization technique to identify corner poses of the surface. Based on the corner poses, a virtual model of the surface is created and matched to the point set to determine the pose of the surface.

Claims

1. A system comprising:

one or more processors;

an imaging sensor;

a projector;

one or more computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform acts comprising;

obtaining a first candidate point set and a second candidate point set using the imaging sensor;

determining a first likelihood that the first candidate point set represents a display medium upon which content is to be projected;

determining a second likelihood that the second candidate point set represents the display medium;

determining that the first candidate point set is a point set that represents the display medium based at least partly on at least one of the first likelihood or the second likelihood;

identifying, from the point set, an observed contour corresponding to a periphery of the display medium;

comparing one or more corner models to the observed contour to identify one or more corner poses corresponding to at least one corner of the display medium;

estimating a size of the display medium based at least in part on the one or more corner poses corresponding to the at least one corner of the display medium;

identifying a pose of the display medium at a first time based at least in part on the size of the display medium;

determining a prior pose of the display medium at a second time that is prior to the first time;

determining a difference between the pose of the display medium and the prior pose of the display medium;

predicting a future pose of the display medium based at least in part on the difference; and

projecting the content onto the display medium from the projector at the pose of the display medium or the future pose of the display medium.

2. The system of claim 1, wherein identifying the pose of the display medium comprises:

creating a model contour of the display medium based at least in part on the size of the display medium; and

aligning the model contour of the display medium to the observed contour.

3. The system of claim 2, wherein aligning the model contour comprises determining a position of the model contour that minimizes differences between points of the model contour and points of the observed contour.

4. The system of claim 1, wherein comparing the one or more corner models comprises determining positions of the corner models that minimize differences between points of the corner models and the points of the observed contour.

5. The system of claim 1, wherein comparing the one or more corner models uses iterative closest point minimization to determine positions of the corner models that minimize differences between points of the corner models and the points of the observed contour.

6. The system of claim 1, the acts further comprising estimating corners of the observed contour prior to comparing the one or more corner models.

7. The system of claim 1, the acts further comprising:

estimating a center of the observed contour; and

identifying corners of the observed contour prior to comparing the one or more corner models, based at least in part on distances of observed contour points from the center of the observed contour.

21. The system of claim 1, wherein the determining that the point represents the display medium is based at least in part on characteristics of the object, the characteristics including at least one of a size of the object, a shape of the object, a color of the object, or an aspect ratio of the object.

22. The system of claim 1, wherein predicting the future pose of the display medium comprises calculating one or more parameters of a motion model that represents at least one of past motion of the display medium or future motion of the display medium.

8. A method comprising:

determining a size of an object;

obtaining a point set;

determining a likelihood that the point set represents at least a portion of the object;

determining that the point set represents the at least the portion of the object based at least partly on the likelihood;

identifying, from the point set, a first contour representing a periphery of the object;

creating a second contour of the object based at least partly on the size of the object;

determining a position of the object at a first time by aligning the second contour with the first contour;

determining a prior position of the object at a second time that is prior to the first time;

determining a difference between the position of the object and the prior position of the object; and

determining a future position of the object based at least in part on the difference.

9. The method of claim 8, further comprising:

determining corners of the first contour;

wherein the determining the position of the object is based at least in part on the corners of the first contour.

10. The method of claim 8, wherein the aligning includes minimizing differences between points of the second contour and the point set of the first contour.

11. The method of claim 10, further comprising determining the position of the object based at least partly on one or more edge models, wherein the one or more edge models comprise corner models.

12. The method of claim 10, determining positions of one or more edge models that minimize differences between the one or more edge models and the point set of the first contour.

13. The method of claim 12, wherein the one or more edge models comprise corner models, the method further comprising estimating corners of the first contour prior to comparing the one or more edge models.

14. The method of claim 12, wherein the one or more edge models comprise corner models, the method further comprising:

estimating a center of the first contour; and

identifying corners of the first contour prior to comparing the one or more edge models, based at least in part on distances of contour points from the center of the first contour.

15. The method of claim 13, wherein comparing the one or more edge models uses iterative closest point minimization to determine positions of the one or more edge models that minimize differences between the one or more edge models and the point set of the first contour.

23. The method of claim 8, wherein predicting the future position of the object comprises calculating one or more parameters of a motion model that represents at least one of past motion of the object or future motion of the object.

16. One or more non-transitory computer-readable media storing computer-executable instructions that, when executed by one or more processors, cause the one or more processors to perform acts comprising:

obtaining a point set;

determining a likelihood that the point set represents at least a portion of a display medium, the display medium having a size;

determining that the point set represents the at least the portion of the display medium based at least partly on the likelihood;

identifying, from the point set, a first contour representing a periphery of the display medium;

creating a second contour corresponding to the size of the display medium;

determining a pose of the display medium at a first time by aligning the second contour with the first contour;

determining a prior pose of the display medium at a second time that is prior to the first time;

determining a difference between the pose of the display medium and the prior pose of the display medium; and

determining a future pose of the display medium based at least in part on the difference.

17. The one or more non-transitory computer-readable media of claim 16, further comprising aligning one or more feature models to the first contour to determine the size of the display medium.

18. The one or more non-transitory computer-readable media of claim 17, wherein aligning the one or more feature models comprises determining positions of the one or more feature models that minimize differences between points of the one or more feature models and the point set of the first contour.

19. The one or more non-transitory computer-readable media of claim 17, wherein aligning the one or more feature models uses iterative closest matching to determine positions of the one or more feature models that minimize differences between points of the one or more feature models and the point set of the first contour.

20. The one or more non-transitory computer-readable media of claim 17, wherein the one or more feature models comprise right-angle corner models.