Hierarchical RBF
In computer graphics, a hierarchical RBF is an interpolation method based on Radial basis functions (RBF). Hierarchical RBF interpolation has applications in the construction of shape models in 3D computer graphics (see Stanford Bunny image below), treatment of results from a 3D scanner, terrain reconstruction, and others.
This problem is informally named as "large scattered data point set interpolation."
The steps of the method (for example in 3D) consist of the following:
- Let the scattered points be presented as set [math]\displaystyle{ \mathbf{P}=\{\mathbf{c}_i=(\mathbf{x}_i,\mathbf{y}_i,\mathbf{z}_i)\vert^{N}_{i=1} \subset \mathbb{R}^3\} }[/math]
- Let there exist a set of values of some function in scattered points [math]\displaystyle{ \mathbf{H}=\{\mathbf{h}_i \vert^{N}_{i=1}\subset \mathbb{R}\} }[/math]
- Find a function [math]\displaystyle{ \mathbf{f}(\mathbf{x}) }[/math] that will meet the condition [math]\displaystyle{ \mathbf{f}(\mathbf{x})=1 }[/math] for points lying on the shape and [math]\displaystyle{ \mathbf{f}(\mathbf{x})\neq1 }[/math] for points not lying on the shape
- As J. C. Carr et al. showed,[1] this function looks like [math]\displaystyle{ \mathbf{f}(\mathbf{x})=\sum_{i=1}^N \lambda_i \varphi(\mathbf{x},\mathbf{c}_i) }[/math] where:
[math]\displaystyle{ \varphi }[/math] — is RBF; [math]\displaystyle{ \lambda }[/math] — is coefficients that are the solution of the system shown in the picture:
For determination of surface, it is necessary to estimate the value of function [math]\displaystyle{ \mathbf{f}(\mathbf{x}) }[/math] in interesting points x. A lack of such method is a considerable complication [2] [math]\displaystyle{ \mathbf{O}(\mathbf{n}^2) }[/math] to calculate RBF, solve system, and determine surface.
Other methods
- Reduce interpolation centers ([math]\displaystyle{ \mathbf{O}(\mathbf{n}^2) }[/math] to calculate RBF and solve system, [math]\displaystyle{ \mathbf{O}(\mathbf{m}\mathbf{n}) }[/math] to determine surface)
- Compactly support RBF ([math]\displaystyle{ \mathbf{O}(\mathbf{n}\log{\mathbf{n}}) }[/math] to calculate RBF, [math]\displaystyle{ \mathbf{O}(\mathbf{n}^{1.2..1.5}) }[/math] to solve system, [math]\displaystyle{ \mathbf{O}(\mathbf{m}\log{\mathbf{n}}) }[/math] to determine surface)
- FMM ([math]\displaystyle{ \mathbf{O}(\mathbf{n}^2) }[/math] to calculate RBF, [math]\displaystyle{ \mathbf{O}(\mathbf{n}\log{\mathbf{n}}) }[/math] to solve system, [math]\displaystyle{ \mathbf{O}(\mathbf{m}+\mathbf{n}\log{\mathbf{n}}) }[/math] to determine surface)
Hierarchical algorithm
An idea of hierarchical algorithm is an acceleration of calculations due to decomposition of intricate problems on the great number of simple (see picture). 
In this case, hierarchical division of space contains points on elementary parts, and the system of small dimension solves for each. The calculation of surface in this case is taken to the hierarchical (on the basis of tree-structure) calculation of interpolant. A method for a 2D case is offered by Pouderoux J. et al.[3] For a 3D case, a method is used in the tasks of 3D graphics by W. Qiang et al.[4] and modified by Babkov V.[5]
References
- ↑ Carr, J.C.; Beatson, R.K.; Cherrie, J.B.; Mitchell, T.J.; Fright, W.R.; McCallum B.C.; Evans, T.R. (2001), “Reconstruction and Representation of 3D Objects with Radial Basis Functions” ACM SIGGRAPH 2001, Los Angeles, CA, P. 67–76.
- ↑ Bashkov, E.A.; Babkov, V.S. (2008) “Research of RBF-algorithm and his modifications apply possibilities for the construction of shape computer models in medical practice”. Proc Int. Conference "Simulation-2008", Pukhov Institute for Modelling in Energy Engineering, [1] (in Russian)
- ↑ Pouderoux, J. et al. (2004), “Adaptive hierarchical RBF interpolation for creating smooth digital elevathion models”, Proc. 12-th ACM Int. Symp. Advances in Geographical information Systems 2004, ACP Press, P. 232–240
- ↑ Qiang, W.; Pan, Z.; Chun, C.; Jiajun, B. (2007), “Surface rendering for parallel slice of contours from medical imaging”, Computing in science & engineering, 9(1), January–February 2007, P 32–37
- ↑ Babkov, V.S. (2008) “Modification of hierarchical RBF method for 3D-modelling based on laser scan result”. Proc. Int. Conference “Modern problems and achievement of radio, communication and informatics”, Zaporizhzhya National Technical University, [2] (in Ukrainian)
 |  |
