A comprehensive system to locate and track objects in two or three
dimensional space, using non-contact video sensing techniques is described. The
need exists to be able to quantify range and proximity of objects that would be
difficult or impossible to measure using standard contact based sensor technology.
Available video technology is surveyed and classified. Then, a hardware system
is assembled that fulfills the project goal, within given budgetary constraints. The
individual components of the system are described in detail.
The theoretical solution for single camera, 2-D positioning, is developed.
A device dependent computer algorithm is developed to perform the object
location. An accurate multi-camera, 3-D positioning algorithm is also developed.
A method to calibrate the cameras is also described and applied. Computer
algorithms to perform calibration and solve the multiple view, 3-D location
geometry are presented. The theoretical equations and most of the algorithms
are transferable, not hardware specific.
Examples using the 2-D model are presented. The first test is a submerged,
single degree of freedom model that was subjected to wave action. Video
tracking data is compared with available positioning data from string
potentiometers. The second test is a surface float application where contact
sensing methods were not possible.
The 3-D algorithm is demonstrated in an above water test. The
longitudinal motion of a linear constrained target is measured with a string
potentiometer and compared with a two-camera, 3-D video interpretation of the
motion. The calibration method is verified with the 3-D algorithm. / Graduation date: 1994
| Identifer | oai:union.ndltd.org:ORGSU/oai:ir.library.oregonstate.edu:1957/36764 |
| Date | 17 December 1993 |
| Creators | Jansky, L. Andrew |
| Contributors | Sollitt, Charles K. |
| Source Sets | Oregon State University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
Page generated in 0.0018 seconds