Improving non-uniform rational B-splines' knot removal with particle swarm optimization

Abstract

Data reduction and shape accuracy are two things that come in mind when it comes to computer graphics research. One such algorithm used is Non-Uniform Rational B-splines (NURBS). The method used to reduce data in NURBS is via knot removal, which relies on an error tolerance value. This paper proposes integrating Particle Swarm Optimization (PSO) to determine the error tolerance value. The approach has made determining the error tolerance value an automatic process. There is a small amount of success in implementing this method thus far. It is not a perfect solution as of yet, but it has potential to unite artificial intelligence algorithms and computer graphics in efforts to promote more avenues of research in this field.