Endoscopic Camera Modeling
The borescope is an optical device used in industry and medicine for the observation of cavities with difficult or limited access. The device consists of a rigid tube, with an eyepiece on one end and an objective lens on the other, linked by a relay optical system. The relay works as an optical repeater, moving the projection center from the objective lens to a virtual point at the back end of the rigid tube. The eyepiece is often ?tted with a coupler lens to enable the use of a CCD camera for digital image acquisition. In medicine the borescope is usually referred as rigid endoscope. Henceforth, we will use the name endoscope instead of borescope.
The endoscope is a non-conventional camera that presents severe non-linear distortions. Moreover it does not necessarily follow perspective projection, specially when mechanical torsion is applied to the probe. In order to study these aspects we performed a ray-based calibration procedure, with no prior assumptions about centrality or specific distortion models.
The accuracy of the achieved ray-based calibration was slightly more than one tenth of millimeter at a depth range up to 40mm. Moreover, the endoscopic camera seems to be well approximated by the central projection model, even when the scope is under strong mechanical torsion. However, performing camera calibration assuming the standard Bouguet toolbox not always leads to the best results. According to our experiments the radial distortion model that better describes the lens distortion in endoscopes is the so called "Division Model".
- [C8] J. P. Barreto, J. M. Santos, P. Menezes and F. Fonseca, Ray-based Calibration of Rigid Medical Endoscopes, Int. Work. on Omnidirectional Vision, Camera Networks and Non-Classical Cameras. Marseille, October 2008.
- [C9] Joao P. Barreto, Rahul Swaminathan and Jose Roquette. Non Parametric Distorion Correction in Endoscopic Medical Images. 3DTV-CON The True Vision, Capture, Transmission and Display of 3D Video. Kos, Greece, May 2007
The presented work was partially/totally funded by The Portuguese Agency for Innovation (ADI) through the QREN co-promotion project n 2009/003512 sponsored by the Operational Program "Factores de Competitividade" and the 10 European Fund for Regional Development.