Paper
15 April 2008 Enhanced optical tracking
Author Affiliations +
Abstract
Enhanced tracking is accomplished by increasing the resolution, frame rate and processing capabilities in tracking dynamic regions of interest for vision applications. In many proven algorithms, the ability to distinguish an object and track it is dependent on the system performance in more than one attribute. We have conducted studies on proven techniques such as Active Appearance Models, Principle Component Analysis and Eigen tracking. All perform better as the camera resolution increases, and camera frame rate increases. Additional opportunities have been observed by combining these techniques, taking advantage of Multicore CPUs, and GPU graphic card processing. Results from an 8 Megapixel commercial sensor combined with a Field Programmable Gate array are presented, and algorithm performance compared with down scaled images of the same scenes, and simulated typical 30 hertz frame rates verses the 120 hertz to 300 hertz typical of this smart camera.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Tracy McSheery "Enhanced optical tracking", Proc. SPIE 6957, Enhanced and Synthetic Vision 2008, 69570A (15 April 2008); https://doi.org/10.1117/12.779441
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Cameras

Image processing

Sensors

Computer vision technology

Machine vision

CMOS sensors

Detection and tracking algorithms

RELATED CONTENT

AKITA: Application Knowledge Interface to Algorithms
Proceedings of SPIE (June 03 2013)
The XH map algorithm A method to process stereo...
Proceedings of SPIE (October 24 2005)
Vision-based natural feature tracking for augmented reality
Proceedings of SPIE (September 05 2002)
Real-time landmark-based optical vehicle self-location
Proceedings of SPIE (February 26 1997)

Back to Top