Title :
Precise High Speed Multi-Target Multi-Sensor Local Positioning System
Author :
Eichel, Justin A. ; Clausi, David A. ; Fieguth, Paul
Author_Institution :
Vision & Image Process. Lab., Univ. of Waterloo, Waterloo, ON, Canada
Abstract :
When used for tracking, the combination of infrared (IR) and an internal measurement unit (IMU) allows researchers and industry to locate objects to within 1 cm at over 200 Hz with a latency less than 2 ms. This novel tracking system uses multiple cameras to triangulate an IR LED placed on the object and utilizes IMU data to measure the object´s orientation and allow the system to be robust against visual occlusions. The cost-effective IR system is robust against illumination allowing the object to be tracked at over 50 feet from the camera positions. This distance is expected to increase with additional development. This paper describes the preliminary algorithms that utilize the information from both the IR and IMU systems in order to precisely track an object. The algorithms are tested against simulated data generated from profiling the hardware and real data collected from a system prototype.
Keywords :
image sensors; object tracking; sensor fusion; IMU data; IR LED; IR system; internal measurement unit; multiple cameras; multitarget multisensor local positioning system; object location; object orientation measurement; tracking system; Acceleration; Cameras; Current measurement; Kalman filters; Position measurement; Radar tracking; Robot sensing systems; Kalman filter; infrared; motion capture; sensor fusion; tracking;
Conference_Titel :
Computer and Robot Vision (CRV), 2011 Canadian Conference on
Conference_Location :
St. Johns, NL
Print_ISBN :
978-1-61284-430-5
Electronic_ISBN :
978-0-7695-4362-8
DOI :
10.1109/CRV.2011.59