Conic Geometry and Autocalibration from Two Images
We show how the classical theory of projective conics provides new insights and results on the problem of 3D reconstruction from two images taken with uncalibrated cameras. The close relationship between Kruppa equations and Poncelet’s Porism is investigated, leading, in particular, to a closed-form geometrically meaningful parameterization of the set of Euclidean reconstructions compatible with two images taken with cameras with constant intrinsic parameters and known pixel shape. An experiment with real images, showing the applicability of the method, is included.
KeywordsCamera autocalibration Conic geometry Kruppa configuration Poncelet’s Porism
Unable to display preview. Download preview PDF.
- 1.Bougnoux, S.: From projective to Euclidean space under any practical situation, a criticism of self-calibration. In: Sixth International Conference on Computer Vision, pp. 790–796, 1998 Google Scholar
- 2.Hartley, R.: Estimation of Relative Camera Positions for Uncalibrated Cameras. Lecture Notes In Computer Science, vol. 588. Proceedings of the Second European Conference on Computer Vision, pp. 579–587, 1992 Google Scholar
- 7.Newsam, G.N., Huynh, D.Q., Brooks, M.J., Pan, H.P.: Recovering unknown focal lengths in self-calibration: an essentially linear algorithm and degenerate configurations. In: ISPRS-Congress XXXI (B3), pp. 575–580, 1996 Google Scholar
- 8.Nister, D., Schaffalitzky, F.: What do four Points in two calibrated images tell us about the epipoles? In: Pajdla, T., Matas, J. (Eds.) ECCV 2004, LNCS 3022, pp. 41–57, Springer Google Scholar
- 9.Ponce, J., McHenry, K., Papadopoulo, T., Teillaud, M., Triggs, B.: On the absolute quadratic complex and its application to autocalibration. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition, vol. I, pp. 780–787, San Diego, CA, June 2005 Google Scholar