Detecting and tracking objects in images

• US 20060188849A1
• Filed: 01/06/2006
• Published: 08/24/2006
• Est. Priority Date: 01/07/2005
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: GestureTek Incorporated (Winning Brands Corp.)
• Inventors: Shamaie, Atid
• Granted Patent: US 7,853,041 B2
• Time in Patent Office: Days
• US Class Current: 434/85
• CPC Class Codes: G06F 18/2135 based on approximation criteria, e.g. principal component analysis
• CPC Class Codes: G06F 18/2413 based on distances to training or reference patterns
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects
• CPC Class Codes: G06V 10/7715 Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods
• CPC Class Codes: G06V 40/107 Static hand or arm
• CPC Class Codes: G06V 40/28 Recognition of hand or arm movements, e.g. recognition of deaf sign language static hand signs G06V40/113

Abstract

According to one disclosed method, coordinates in a multi-dimensional space are determined for an image point characterizing a particular object. An equation describing a model in the multi-dimensional space is provided. The model is characteristic of a set of training images of one or more other objects. The coordinates are applied to the equation to determine a distance between the image point and the model. Based on the determined distance, a determination is made as to whether the particular object matches the one or more other objects. A set of training images may be received. A multi-dimensional space (e.g., eigenspace) may be determined based on the set of training images. A set of training points may be generated by projecting the set of training images into the multi-dimensional space. An equation describing a model in the multi-dimensional space that is characteristic of the set of training points may be determined.

Claims

1. A method comprising:

determining coordinates for an image point in a multi-dimensional space, the image point characterizing a particular object;

providing an equation describing a model in the multi-dimensional space, the model being characteristic of a set of training images of one or more other objects;

applying the coordinates to the equation to determine a distance between the image point and the model; and

determining whether the particular object matches the one or more other objects based on the determined distance.

2. The method of claim 1 further comprising:

receiving an image of the particular object, the image having a number of data elements, and the multi-dimensional space having a dimensionality that is lower than the number of data elements; and

wherein determining coordinates comprises projecting the image into the multi-dimensional space to produce the coordinates for the image point in the multi-dimensional space.

3. The method of claim 2 wherein the model is a model of a set of training points in the multi-dimensional space, each of the training points in the set corresponding to one or more images in the sequence of training images.

4. The method of claim 2 wherein the image of the particular object is normalized before being received.

5. The method of claim 4 wherein normalizing the image of the particular object accounts for variations in brightness.

6. The method of claim 4 wherein normalizing the image of the object comprises applying a histogram equalization technique to the image of the particular object.

7. The method of claim 1 wherein the equation describes a hyperboloid.

8. The method of claim 1 wherein the equation describes a cone.

9. The method of claim 1 wherein the equation describes a line.

10. The method of claim 1 wherein the equation describes a combination of a cone and a line.

11. The method of claim 1 wherein the particular object includes an object from the set consisting of a finger, a pen, and substantially cylindrical objects.

12. The method of claim 1 wherein providing the equation comprises selecting the equation from among a first equation and a second equation, the first equation describing a first model characteristic of a first set of training images, and the second equation describing a second model characteristic of a second set of training images.

13. A method comprising:

receiving an image of a particular object, the image having a number of data points;

projecting the image into a multi-dimensional space having a dimensionality that is lower than the number of data points to produce coordinates for an image point in the multi-dimensional space, the image point characterizing the particular object;

providing an equation describing a model in the multi-dimensional space, the model being a model for a set of training points in the multi-dimensional space, each of the training points in the set corresponding to one or more images in a training set of images of one or more other objects;

applying the coordinates to the equation to determine a distance between the image point and the model; and

determining whether the particular object matches the other objects based on the determined distance.

14. The method of claim 13 wherein providing the equation comprises selecting the equation from among a first equation and a second equation, the first equation describing a first model characteristic of a first set of training images, and the second equation describing a second model characteristic of a second set of training images.

15. A system comprising:

a camera; and

a processing device coupled to the camera and configured to;

determine coordinates for an image point in a multi-dimensional space, the image point characterizing a particular object;

provide an equation describing a model in the multi-dimensional space, the model being characteristic of a set of training images of one or more other objects;

apply the coordinates to the equation to determine a distance between the image point and the model; and

determine whether the particular object matches the one or more other objects based on the determined distance.

16. The system of claim 15 wherein the camera comprises an IR camera.

17. The system of claim 16 further comprising an IR source.

18. The system of claim 15 further comprising a screen disposed in front of the camera.

19. The system of claim 18 wherein the screen is at least translucent such that light reflected by an object disposed on a side of the screen opposite the camera can pass through the screen and be received by the camera.

20. A method comprising:

providing a tool for

determining coordinates for a particular image point in a multi-dimensional space, the particular image point characterizing a particular object;

providing an equation describing a model in the multi-dimensional space, the model being characteristic of a set of training images of one or more other objects;

applying the coordinates to the equation to determine a distance between the particular image point and the model; and

determining whether the particular object matches the one or more other objects based on the determined distance.

21. The method of claim 20 further comprising:

receiving the set of training images, each of the images in the set having a number of data elements;

determining, based on the set of training images, the multi-dimensional space having a dimensionality that is lower than the number of data elements;

generating a set of training points by projecting each image in the set of training images into the multi-dimensional space to produce coordinates for a corresponding training point in the multi-dimensional space, each training point in the set of training points corresponding to at least one of the images in the set of training images; and

determining the equation describing the model in the multi-dimensional space characteristic of the set of training images.

22. The method of claim 21 wherein each image in the set of training images is normalized before being received.

23. The method of claim 22 wherein normalizing each image in the set of training images accounts for variations in brightness.

24. The method of claim 23 wherein normalizing each image in the set of training images comprises applying a histogram equalization technique to each image in the set of training images.