Share Email Print

Proceedings Paper

Real-time videogrammetry uncertainty computation and test results
Author(s): Scott C. Sandwith
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

This report documents system uncertainty, best practices, and process limitations for the real-time videogrammetry system developed by Metronor, ASA. The report describes the system and the tests used to establish its performance. It then documents the test result and draws conclusions. An uncertainty of 23.6 ppm within the 3.2 by 2.8 by 1.5-meter test volume was achieved with the baseline between the cameras parallel to the length of the target field. A slightly lower uncertainty of 18.5 ppm was achieved when the cameras were setup with their baseline normal to the length of the target field.Results suggest that the orientation of the camera baseline relative to the target field does not play a significant role when targets are more than 2 meters apart. Testing shows a significant dependence in system accuracy relative to the apex angle between the cameras and the targets. When the apex angle was between 60 degrees and 95 degrees the performance specification was met. Conclusions drawn from a series of two-sample F-tests of RSS error values establish that sampling a point more than three times does not significantly reduce its uncertainty. This result is in part explained by the redundant fitting of the five-diode points on the light pen: the fitting effects an averaging process even on samples of one.

Paper Details

Date Published: 29 December 1998
PDF: 12 pages
Proc. SPIE 3520, Three-Dimensional Imaging, Optical Metrology, and Inspection IV, (29 December 1998); doi: 10.1117/12.334325
Show Author Affiliations
Scott C. Sandwith, Boeing Commercial Airplane Group (United States)

Published in SPIE Proceedings Vol. 3520:
Three-Dimensional Imaging, Optical Metrology, and Inspection IV
Kevin G. Harding; Donald J. Svetkoff; Katherine Creath; James S. Harris, Editor(s)

© SPIE. Terms of Use
Back to Top