![](bg8.png)
5 CONCLUSION
We have developed a new automatic model-based
face recognition system, which includes both non-
iterative registration and the representation of 3D
face models by shape descriptors. By registering
point clouds to a canonical position, we overcome
the pose-variation problem. Unlike ICP algorithm,
this non-iterative registration process is scale
invariant. An efficient B-spline surface-fitting
technique is developed to reconstruct underlying
surface for the registered data set. A new knot vector
standardisation technique is proposed to allow a
direct one-to-one mapping relationship from the
object space to a parameter space. Subsequently, a
compact parametric representation of 3D objects is
obtained. The system has been tested on a personal
computer (Pentium 4/512M RAM). Compared with
the existing method of closed surface
parameterisation, which takes 33s to 536s depending
on the complexity of the objects [BGK96], the
registration and modelling process introduced in this
paper only takes 2 seconds, on an average sized
point cloud (about 25,000 vertices).
Although surface distance can be used as a
metric for face recognition, it may not be very
sufficient since no explicit geometric information is
employed. Our future work is to integrate geometric
information into recognition methods. For example,
we may turn a recognition problem into a
classification problem of the shape descriptors.
With the proposed surface representation, it is
possible to analyse facial component separately. As
the geometry of B-spline surface can be inferred
from the shape descriptors, we can delineate facial
areas, e.g. forehead, nose, mouth, chin, from the
parameter space, and weigh each part separately in
the recognition metric to reduce the influence of
facial expression. The areas potentially affected by
facial expression will be given lower weight.
REFERENCES
Besl, P.J., and McKay, N.D. (1992) A Method for
Registration of 3D Shapes. IEEE Pattern Analysis and
Machine Intelligence, vol. 14, No. 2, 239-256.
Chen, Y., and Medioni, G. (1992) Object modelling by
registration of multiple range images. Image and
Vision Computing, vol. 10, No. 3, 145-155.
Campbell, R., Flynn, P. (2001) A Survey of Free-form
Object Representation and Recognition Techniques,
Computer Vision and Image Understanding, vol. 81,
pp. 166-210.
Eck, M., DeRose, T., Duchamp, T., Hoppe, H.,
Lounsbery, M., and Stuetzle, W. (1995)
Multiresolution Analysis of Arbitrary Meshes. In
Computer Graphics (Proceedings of SIGGRAPH’ 95),
pages 173-182.
Eck, M., and Hoppe, H. (1996) Automatic Reconstruction
of B-Spine Surfaces of Arbitrary Topological Type.
Proc. 23rd Int'l. Conf. on Computer Graphics and
Interactive Techniques SIGGRAPH '96, ACM, New
York, NY. pp. 325-334.
Fusiello, A., Castellani, U., Ronchetti, L., and Murino, V.
(2002) Model Acquisition by Registration of Multiple
Acoustic Range Views, Computer Vision, ECCV2002,
Springer, pp. 805-819.
Flusser, J. and Zitova, B. (2003) Image Registration
Methods: A Survey, Image and Vision Computing,
vol. 21, pp. 977-1000.
Godin, G. and Boulanger, P. (1995) Range Image
Registration Through Viewpoint Invariant
Computation of Curvature, IAPRS, 30 (5/W1), pp.
170-175.
Godin, G., Rioux, M. and Baribeau, R. (1994) Three-
dimensional Registration Using Range and Intensity
Information, SPIE, vol. 2350, Videometrics III, pp.
279-290.
Godin, G., Laurendeau, D. and Bergevin, R. (2001) A
Method for the Registration of Attributed Range
images, International Conference on 3D Imaging and
Modeling, Quebec, pp. 179-186.
Hoschek, J., Schneider, F. –J. and Wassum, P. (1989)
Optimal Approximate Conversion of Spline Surfaces,
Computer Aided Geometric Design, vol. 6, issue 4, pp.
293-306.
Krishnamurthy, V. and Levoy, M. (1996) Fitting Smooth
Surfaces to Dense Polygon Meshes, ACM-0-89791-
746-4/96/008.
Lele, S. and Richtsmeier, J. (2001) An Invariant Approach
to the Statistical Analysis of Shapes. Boca Raton:
Chapman & Hall/CRC.
Milroy, M., Bradley, C., Vickers, G. and Weir, D. (1995)
G1 Continuity of B-spline Surface Patches in Reverse
Engineering. CAD 27, 471-478.
Rogers, D. and Fog, N. (1989) Constrained B-spline Curve
and Surface Fitting, Computer Aided Design, vol. 21,
pp. 641-648.
Schmitt, F., Barsky, B. and Du, W. H. (1986) An Adaptive
Subdivision Method for Surface Fitting from Sampled
Data. Computer Graphics (SIGGRAPH’ 86
Proceedings), volume 20. pp. 179-188.
Sarkar, B. and Menq, C. (1991) Parameter Optimization in
Approximating Curves and Surfaces to Measurement
Data, Computer Aided Geometric Design, vol. 8, pp.
267-290.
Watt, A. and Watt, M. (1992) Advanced Animation and
Rendering Techniques: Theory and Practise. Addison-
Wesley.
Zhang, Z. (1994) Iterative Point Matching for Registration
of Free-form Curves and Surfaces. International
Journal of Computer Vision, vol. 13, No. 2, pp.119-
152.
VISAPP 2006 - IMAGE UNDERSTANDING
208