Milan Blagojević, PhD student,

Faculty of Mechanical Engineering, Sestre Janjić 6, Kragujevac, blagoje@kg.ac.rs

Miroslav Živković, PhD, Full professor,
Faculty of Mechanical Engineering, Sestre Janjić 6, Kragujevac, zile@kg.ac.rs
Bojana Rosić, PhD student,
Technical University of Braunschweig, Institut fur Wissenschaftliches Rechnen, Hans
Sommer Strasse 65, Braunschweig, Deutschland, bojana.rosic@tu-bs.de

QUALITY 3D SURFACE RECONSTRUCTION

BASED ON POINT CLOUD
GENERATED BY OPTICAL MEASURING TECHNIQUES

ABSTRACT:

Today's products are mostly very complex shapes. The new products or tools development
are often based on already existing products, prototypes or physical models. Such free form
surface can be accurately reconstructed only on the basis of a large number of measuring
points. By the classic measuring methods, it is very difficult or impossible to measure
sufficient number of points. The measuring project described in this paper is more complex,
and the digitization is realized by both TRITOP and ATOS measurement systems. A result
of digitization with optical measuring systems ATOS and TRITOP are detailed poligonized
mesh or point cloud.
Depending on application, creating of the model’s surface is carried out with introduction of
certain assumptions and approximations, based on knowledge of processing technology of
model whose geometry is reconstructed. All commands necessary for reconstruction are
placed in CATIA Workbenches under group Shape (Quick Surface Reconstruction,
Digitized Shape Editor, Generative Shape Design, Free Style, and Automotive Class A).
The presented procedure is hardware- and time-consuming. It generates accurate digitized
object’s surface, mathematically defined by NURBS.

KEYWORDS:

Optical 3D digitization, 3D surface reconstruction, Reverse Engineering, Product

development, CAD/CAM/CAE