UCL logo

UCL Discovery

UCL home » Library Services » Electronic resources » UCL Discovery

Multistation bundle adjustment with a machine vision parallel camera system-an alternative to the perspective case for the measurement of small objects

Rova, M; Robson, S; Cooper, MAR; (2008) Multistation bundle adjustment with a machine vision parallel camera system-an alternative to the perspective case for the measurement of small objects. In: (pp. pp. 45-50).

Full text not available from this repository.

Abstract

Photogrammetric measurement and modelling procedures are dependent on accurate camera calibration and orientation. The accuracy of image-based registration and texture mapping approaches can be deficient up to the underlying camera projection. In the case of small objects an alternative to the conventional central projection model can be to use parallel projection cameras which deliver small volumes under constant sampling. Cameras of variant geometries have been studied by the computer vision community; but they impose algorithmic assumptions into the projective model which can degrade the stability of the solutions. This paper proposes the development of a multiphoto bundle adjustment for parallel cameras based on an analytical mathematical model with quality assessment. The model is treated for multiphoto situations; the image to object space correspondence is established under invariant scaling through the pseudo exterior orientation elements; the camera is calibrated for the first term of the radial lens distortion. Initialization is recovered indirectly through perspective-based strategies and direct solutions based on orthographic projection. The algorithmic stability is verified with true parallel projection image sequences of 3D calibration structures captured with a telecentric camera system within controlled calibration conditions.

Type: Proceedings paper
Title: Multistation bundle adjustment with a machine vision parallel camera system-an alternative to the perspective case for the measurement of small objects
URI: http://discovery.ucl.ac.uk/id/eprint/48888
Downloads since deposit
0Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item