Post on 10-Feb-2016
description
Computer aided geometric designwith Powell-Sabin splinesSpeaker: 周 联
2008.10.29
Ph.D Student Seminar
What is it?
C1-continuous quadratic splines defined on an arbitrary triangulation in Bernstein-Bézier representation
Why use it?
PS-Splines vs. NURBS suited to represent strongly irregular objects
PS-Splines vs. Bézier triangles smoothness
Main works M.J.D. Powell, M.A. Sabin. Piecewise quadratic approximations on triangles. ACM Trans. Math. Softw., 3:
316–325, 1977. P. Dierckx, S.V. Leemput, and T. Vermeire. Algorithms for surface fitting using Powell-Sabin splines, IMA
Journal of Numerical Analysis, 12, 271-299, 1992. K. Willemans, P. Dierckx. Surface fitting using convex Powell-Sabin splines, JCAM, 56, 263-282,1994. P. Dierckx. On calculating normalized Powell-Sabin B-splines. CAGD, 15(1):61–78, 1997. J. Windmolders and P. Dierckx. From PS-splines to NURPS. Proc. of Curve and Surface Fitting, Saint-Ma
lo, 45–54. 1999. E. Vanraes, J. Windmolders, A. Bultheel, and P. Dierckx. Automatic construction of control triangles for s
ubdivided Powel-Sabin splines. CAGD, 21(7):671–682, 2004. J. Maes, A. Bultheel. Modeling sphere-like manifolds with spherical Powell–Sabin B-splines. CAGD, 24 7
9–89, 2007. H. Speleers, P. Dierckx, and S. Vandewalle. Weight control for modelling with NURPS surfaces. CAGD, 2
4(3):179–186, 2007. D. Sbibih, A. Serghini, A. Tijini. Polar forms and quadratic spline quasi-interpolants on Powell–Sabin parti
tions. IMA Applied Numerical Mathematic, 2008. H. Speleers, P. Dierckx, S. Vandewalle. Quasi-hierarchical Powell–Sabin B-splines. CAGD, 2008.
Authors
Professor atKatholieke Universiteit Leuven(鲁汶大学 ), Computerwetenschappen.
Paul Dierckx
Research Interests: Splines functions, Powell-Sabinsplines. Curves and Surface fitting. Computer Aided Geometric Design. Numerical Simulation.
Authors
Stefan Vandewalle
Professor at Katholieke Universiteit Leuven, Faculty of, CS
Research Projects:
Algebraic multigrid for electromagnetics. High frequency oscillatory integrals and integral equations. Stochastic and fuzzy finite element methods. Optimization in Engineering. Multilevel time integration methods.
Problem State (Powell,Sabain,1977)
9 conditions vs. 6 coefficients22),( fveuvducvbuavu
A lemma
PS refinement
Nine degrees of freedom
PS refinement
The dimension equals 3n.
Other refinement
A theorem
Normalized PS-spline(Dierckx, 97)
Local support
Convex partition of unity.
Stability
Obtain the basis function Step 1.
Obtain the basis functionStep 2.
Obtain the basis functionStep 3.
Obtain the basis function
Step 4.
PS-splines
Choice of PS triangles
To calculate triangles of minimal area
Simplify the treatment of boundary conditions
PS control triangles
PS control triangles
A Bernstein-Bézier representation
A Powell-Sabin surface
Local support(Dierckx,92)
Explicit expression for PS-splines
Normalized PS B-splines Necessary and sufficient conditions:
The control points
The control points
The Bézier ordinates of a PS-spline
Spline subdivision(Vanraes, 2004) Refinement rules of the triangulation
Refinement rules
Construction of refined control triangles
Triadically subdivided spline
Application
Visualization
QHPS(Speleers,08)
Data fitting
Data fitting
Rational Powell-Sabin surfaces
B-spline representation for PS splines on the sphere(Maes,07)
Thank you!