The calculation of parametric NURBS surface interval values using neural networks

  • Authors:
  • Erkan Ülker;Ahmet Arslan

  • Affiliations:
  • Computer Engineering Department, Selçuk University, Konya, Turkey;Computer Engineering Department, Selçuk University, Konya, Turkey

  • Venue:
  • ICCS'06 Proceedings of the 6th international conference on Computational Science - Volume Part II
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Three dimensional coordinate values of parametric NURBS (Non-Uniform Rational B-Splines) surfaces are obtained from two dimensional parameters u and v. An approach for generating surfaces produces a model by giving a fixed increase to u and v values. However, the ratio of three dimensional parameters increases and fixed increase of u and v values is not always the same. This difference of ratio costs unrequired sized breaks. In this study an artificial neural network method for simulation of a NURBS surface is proposed. Free shaped NURBS surfaces and various three dimensional object simulations with different patches can be produced using a method projected as network training with respect to coordinates which are found from interval scaled parameters. Experimental results show that this method in imaging modeled surface can be used as a simulator.