| |
Project
descriptions
During his visit, Miklós
has been working on the following projects:
- Turbine blade section curve
design
- The purpose of this project is to
give an efficient and easy-to-use method to create and modify section curves
given by their endpoints and tangent-functions. Based on a double LSQ method,
the developed technique gives the possibility of the modification of the
B-spline section curves by direct alteration of the tangent-functions, which
are B-spline arcs as well.
- Scattered data modeling by
B-spline surfaces using artificial neural networks
- The approximation of unorganized
points by B-spline surfaces is a well-known problem and several types of
methods have been developed in the past few years. Our approach applies an
artificial neural network, called Kohonen self-organizing map, to order the
data in a preprocessing step, the result of which is the quadrilateral control
mesh of the future surface. The method was also applied for local update of
surfaces as well as special (ruled, developable) surface
modeling.
- Geometrical aspects of knot
modification of B-spline and NURBS curves and surfaces
- These curves and surfaces play
central role in CAGD, hence the modification of them is also of great
importance, where all of the known methods use control points and/or weights
for shape control. The aim of this project is to describe the theoretical
(geometrical) background of knot modification of these curves and surfaces and
develope constraint-based shape control methods using knot alteration
exclusively, or together with the modification of other
data.
Publications
The following publications
summarize some of the results of the research work that Miklós has participated
in during his stay:
- The effect of knot modifications
for the shape of B-spline curves (with I. Juhasz) , Journal for Geometry
and Graphics 5, 111-119 (2001)
- Developable surface modeling by
Kohonen neural network (with E. Kovacs), Computers and Mathematics with
Applications (accepted for publication)
- B-spline surface from scattered
data by Kohonen neural network (submitted)
- Local update of B-spline surfaces
by Kohonen neural network (submitted)
Conferences
|
