A new algorithm for generating quadrilateral meshes and its application to FE-based image registration

Proceedings, 12th International Meshing Roundtable, Sandia National Laboratories, pp.159-170, Sept. 2003

MESHING
RESEARCH
CORNER

12th International Meshing Roundtable
September 14-17, 2003
Santa Fe, New Mexico, U.S.A.

Rutgers University, Camden, NJ 08102, USA
rsuneeta@camden.rutgers.edu

University of Pennsylvania, Philadelphia, PA 19104, USA
{marcelos, tessa}@seas.upenn.edu

Universidade Federal de Mato Grosso do Sul, Campo Grande, MS 79070-900, Brazil

Abstract
The use of finite element (FE) analysis in the simulation of physical phenomenon over the human body has necessitated the construction of meshes from images. Despite the availability of several tools for generating meshes for FE-based applications, most cannot deal directly with the raw pixel-wise representation of image data. Additionally, some are optimized for the construction of much simpler shapes than those encountered with the human body. In this work, we introduce a new algorithm to obtain strictly convex quadrilateral meshes of bounded size from triangulations of polygonal regions with or without polygonal holes. We present an approach to construct quadrilateral meshes from segmented images from the aforementioned algorithm, and a quantitative analysis of the quality of the meshes generated with our algorithm with respect to the performance of a FE-based image registration method that takes image meshes as input.

Download Full Paper (PDF Format)