Technical Briefs

    Geometric Modeling

    Thursday, 21 November

    14:15 - 16:00

    Room S223

    3D Reconstruction from Drawings with Straight and Curved Edges

    This paper presents a method to recover solid objects from 2D drawings with both straight and curved edges. A curved edge is approximated by two straight lines, and the resulting straight line drawing is recovered as a polyhedron using an established method, and the curved object recovered from this polyhedron.


    Lee Yong Tsui,Nanyang Technological University
    Fen Fang, Nanyang Technological University

    Non-Convex Hull Surfaces

    The NCH Signed Distance function is constructed from oriented point cloud and evaluated on the vertices of a volumetric mesh, regular or adaptive, and an isosurface algorithm is used to generate a polygonal Non-Convex Hull (NCH) mesh. Despite its simplicity, this simple algorithm produces high quality polygon meshes..


    Gabriel Taubin, Brown University

    Progressive Medial Axis Filtration

    We propose a progressive medial axis simplification method inspired from surface optimization techniques which retains the geometric intuition of the scale axis transform, and computes a hierarchy of simplified medial axes by means of successive edge-collapses that can be browsed in real-time.

    We propose a progressive medial axis simplification method inspired from surface optimization techniques which retains the geometric intuition of the scale axis transform, and computes a hierarchy of simplified medial axes by means of successive edge-collapses that can be browsed in real-time.


    Noura Faraj, Telecom ParisTech - CNRS/LTCI
    Jean-Marc Thiery, Telecom ParisTech - CNRS/LTCI
    Tamy Boubekeur, Telecom ParisTech - CNRS/LTCI

    Generating Multi-style Paper Pop-up Designs using 3D Primitive Fitting

    This paper presents an automatic algorithm and the underlying theory for producing paper pop-up designs from 3D models. We incorporate different pop-up styles by using a mesh abstraction technique that fits volumetric primitives unto a 3D mesh, which are later mapped to basic pop-up structures.


    Conrado Jr Ruiz, National University of Singapore
    Sang Le, National University of Singapore
    Kok-Lim Low, National University of Singapore

    Cross-sectional Structual Analysis for 3D Printing Optimization

    We propose a novel cross-sectional structural analysis technique that efficiently detects critical stress inside a 3D object. Unlike finite element methods, our method does not require a volumetric mesh, enabling interactive speed. The orientation of an object is optimized to increase mechanical strength when manufactured with 3D printing.


    Nobuyuki Umetani, Autodesk Research
    Ryan Schmidt, Autodesk Research