Face Recognition From Range Data by Structural Analysis
A primal approach to the analysis of a human face through its 3-D image is proposed. This analysis is based on local pattern concepts deduced from the intrinsic properties of a surface described by its local curvature characteristics. Some theoretical results about curvature are recalled and then used for the extraction of characteristic geometrical features from the face surface. Recognition can be achieved by using a pattern vector of distances calculated from these features. Experimental results are provided to illustrate the various steps of the approach.
Unable to display preview. Download preview PDF.
- P. Baylou, E. H. Bouyakhf, G. Bousseau & A. Mora, “Analyse Automatique du Profil du Visage. Recherche du Meilleur Classifieur a Fin d’Identification.”, Proc. 3rd AFCET Conf. on Pattern Recognition and Artificial Intelligence, Nancy(France), September 1981, 371–382.Google Scholar
- M.Nagao, “Control Strategies in Pattern Analysis”, Proc. of 6th ICPR, Munich, October 1982, 996–1006.Google Scholar
- M. Brady, J. Ponce, A. Yuille & H. Asada, “Describing Surfaces”, Proc. of 2nd Int. Symp. Rob. Res., H. Hanafusa & H. Inone Eds, MIT Press, Cambridge, 5–16.Google Scholar
- P.Besl & R.Jain, “Intrinsic and Extrinsic Surfaces Characteristics”, Proc. of CVPR, June 1985, San-Francisco, 226–233.Google Scholar
- M. Richetin, P. Saint-Marc & J.T. Lapreste, “ Describing Greylevel Textures through Curvature Primal Sketching”, Proc. of ICASSP, Tokyo, April 1986.Google Scholar
- P. Saint-Marc & M. Richetin, “Structural Filtering from Curvature Information”, Proc. of 1985 Computer Vision and Pattern Recognition, Miami, June 1986.Google Scholar
- P. Besl & R. Jain, “Range Image Understanding”, Proc. of CVPR, San-Francisco, June 1985, 430–449.Google Scholar
- M.E. Mortenson, “Geometric Modeling”, Wiley, New-York, 1985.Google Scholar