Open Collections

A high accuracy CMOS camera based 3D optical tracking system design

University of British Columbia

This thesis presents the design and implementation of a high accuracy optical tracking system. The primary contributions of this thesis include the integration of the 3D optical tracking systems, noise correction algorithms, target location algorithms, and camera calibration algorithms with lens distortion identification. With the best combination of these algorithms, the achieved 3D RMS error for the 3D position estimation is 41 for a measuring range of 500 mm by 500 mm on XY plane and 10 mm in Z. For most machines and robots, MDOF motion sensing in real time is required. Current optical tracking systems for MDOF motion measurement are either low accuracy or low measurement rate. For instance, none of these systems can provide over 10 kHz measurement rate for real time sensing. The optical tracking system presented in this thesis can potentially achieve high measurement rate up to 10 kHz and micron level accuracy. Target location algorithms are widely studied. However, to the best of the author’s knowledge, no experimental comparison of these different algorithms in 3D optical tracking systems was found in the literature. In addition, no report on the accuracy comparison of different lens distortion calibration methods was presented in the literature. Actually, both target location and lens calibration methods have great impact on the accuracy of the 3D optical tracking. Therefore, this thesis also investigates on the performance evaluation of target location algorithms in the 3D tracking system and identifies the lens distortion model for the lenses.

Text

2010-10-12

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/29121

Mechanical Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/