ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

On Using Projective Relations for Calibration and Estimation in a Structured-Light Scanner

Dhillon, Daljit Singh and Govindu, Venn Madhav (2009) On Using Projective Relations for Calibration and Estimation in a Structured-Light Scanner. In: Proceedings of the 5th International Symposium on Advances in Visual Computing: Part, 30-DEC 02, 2009, Las Vegas, pp. 831-842.

Full text not available from this repository. (Request a copy)
Official URL: http://portal.acm.org/citation.cfm?id=1696121.1696...

Abstract

For structured-light scanners, the projective geometry between a projector-camera pair is identical to that of a camera-camera pair. Consequently, in conjunction with calibration, a variety of geometric relations are available for three-dimensional Euclidean reconstruction. In this paper, we use projector-camera epipolar properties and the projective invariance of the cross-ratio to solve for 3D geometry. A key contribution of our approach is the use of homographies induced by reference planes, along with a calibrated camera, resulting in a simple parametric representation for projector and system calibration. Compared to existing solutions that require an elaborate calibration process, our method is simple while ensuring geometric consistency. Our formulation using the invariance of the cross-ratio is also extensible to multiple estimates of 3D geometry that can be analysed in a statistical sense. The performance of our system is demonstrated on some cultural artifacts and geometric surfaces.

Item Type: Conference Paper
Additional Information: Copyright of this article belongs to Portal.
Keywords: 3D Reconstruction,Cross-Ratio,Homography, Projector Calibration,Structured-Light Scanner,System Calibration
Department/Centre: Division of Electrical Sciences > Electrical Engineering
Date Deposited: 12 Jul 2010 07:19
Last Modified: 12 Jul 2010 07:19
URI: http://eprints.iisc.ernet.in/id/eprint/29200

Actions (login required)

View Item View Item