Automatic Calibration of a Single-Projector Catadioptric Display System
Abstract
We describe the calibration of a catadioptric omnidirectional video projection system that adjusts its projection to the geometry of any scene by means of a rotating camera. Correction of geometric distortions requires 3D reconstruction of the scene. A camera is used to detect projected point features and calibration is performed in three successive steps: precalibration of camera assuming pure rotation, precalibration of catadioptric projector under central approximation and calibration of the global system, by minimizing the squared distance between the reflected and perceived rays, and by relaxing previous constraints, to refine values of extrinsic parameters. Simulation is used to validate estimated values of parameters and distance between the 3D reconstruction of the projection room and its expected geometry. Influence of noise in detected point coordinates is studied and preliminary results for the reconstruction and projection in real conditions are reported.
Cite
Text
Astre et al. "Automatic Calibration of a Single-Projector Catadioptric Display System." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2008. doi:10.1109/CVPR.2008.4587788Markdown
[Astre et al. "Automatic Calibration of a Single-Projector Catadioptric Display System." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2008.](https://mlanthology.org/cvpr/2008/astre2008cvpr-automatic/) doi:10.1109/CVPR.2008.4587788BibTeX
@inproceedings{astre2008cvpr-automatic,
title = {{Automatic Calibration of a Single-Projector Catadioptric Display System}},
author = {Astre, Benjamin and Sarry, Laurent and Lohou, Christophe and Zeghers, Eric},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
year = {2008},
doi = {10.1109/CVPR.2008.4587788},
url = {https://mlanthology.org/cvpr/2008/astre2008cvpr-automatic/}
}