Back to results
US 7,227,526 B2Granted
Video-based image control system
AssigneeGestureTek, Inc.
InventorsEvan Hildreth, Francis MacDougall
Life of this patent
Estimating the term…
Abstract
A method of using stereo vision to interface with a computer is provided. The method includes capturing a stereo image, and processing the stereo image to determine position information of an object in the stereo image. The object is controlled by a user. The method also includes communicating the position information to the computer to allow the user to interact with a computer application.
Claims24
A stereo vision system for interfacing with an application program running on a computer, the stereo vision system comprising:
- first and second video cameras arranged in an adjacent configuration and operable to produce at least first and second stereo video images;
- and
- a processor operable to receive the first and second stereo video images and detect objects appearing in an intersecting field of view of the cameras, the processor executing a process to:
- define an object detection region in three-dimensional coordinates relative to a position of the first and second video cameras;
- divide the first and second stereo video images into features;
- pair features of the first stereo video image with features of the second stereo video image;
- generate a depth description map, the depth description map describing the position and disparity of paired features relative to the first and second stereo video images;
- generate a scene description based upon the depth description map, the scene description defining a three-dimensional position for each feature;
- cluster adjacent features;
- crop clustered feature based upon predefined thresholds;
- analyze the three-dimensional position of each clustered feature within the object detection region to determine position information of a control object;
- and
- map the position information of the control object to a position indicator associated with an application program as the control object moves within the object detection region.
10 more elements
A method of using computer vision to interface with a computer, the method comprising:
- capturing at least first and second images of a scene;
- dividing the first and second images into features;
- pairing features of the first image with features of the second image;
- generating a depth description map, the depth description map describing the position and disparity of paired features relative to the first and second images;
- generating a scene description based upon the depth description map, the scene description defining a three-dimensional position for each feature;
- clustering adjacent features;
- cropping clustered features based upon predefined thresholds;
- defining an object detection region;
- analyzing the three-dimensional position of each clustered feature within the object detection region to determine position information of an object;
- and
- using the position information to control a computer application.
8 more elements
Dependencies are inferred from the claim text.