Highly accurate vision based surface reconstruction system

Rahman, Taufiqur (2009) Highly accurate vision based surface reconstruction system. Masters thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (9MB)

Abstract

Photogrammetry has come a long way since cameras were employed for aerial mapping tasks. In this digital era, photogrammetry evolved to be used in close range measurements for industrial purposes, thanks to inexpensive and off-the-shelf digital cameras and superior computing power. But how accurately close range photogrammetry can execute a measurement is a question still unexplored by the researchers. This thesis attempts this question to find a vision based measurement solution that employs conventional sensing means yet produces highly accurate results. In a way, the proverbial "pushing the limit" was the main focus of this work. -- It was found that geometrical calibration of digital cameras is the major task in developing a vision based measurement system. This work proposes a novel calibration technique with an improved approach towards implementation. Experimental results and computer simulation confirmed the accuracy produced by the proposed camera calibration technique. A laser projector emitting a linear pattern was used as an additional sensing element in complement to the camera. Finally a spherical surface was digitized in order to assess the performance of the overall system.

Actions (login required)

View Item