Design of a Chromatic 3D Camera with an End-to-End Performance Model Approach

P. Trouve, F. Champagnat, G. Le Besnerais, G. Druart, J. Idier; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2013, pp. 953-960

Abstract


In this paper we present a new method for the design of a 3D single-lens single-frame passive camera. This camera has a chromatic lens and estimates depth based on a depth from defocus technique (DFD). First we develop an original calculation of the Cram??r Rao Bound to predict the theoretical camera accuracy. This model takes into account the optical parameters through the camera Point Spread Function (PSF) and the algorithms parameters applied to the raw image for depth estimation and image restoration. This model is then used for the end-to-end design of a chromatic camera, dedicated to a small UAV, that is realized and experimentally validated.

Related Material


[pdf]
[bibtex]
@InProceedings{Trouve_2013_CVPR_Workshops,
author = {Trouve, P. and Champagnat, F. and Le Besnerais, G. and Druart, G. and Idier, J.},
title = {Design of a Chromatic 3D Camera with an End-to-End Performance Model Approach},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2013}
}