2007ProceedingsRequires access

Discrete Delaunay: Boundary extraction from voxel objects

Dobrina Boltcheva, Dominique Bechmann, Sylvain Thery

Open publisher page 3 citations

Abstract

We present a discrete approach for boundary extraction from 3D image data. The proposed technique is based on the duality between the Voronoi graph computed accross the digital boundary and the Delaunay triangulation. The originality of the approach is that algorithms perform only integer arithmetic and the method does not suffer from standard round problems and numerical instabilities in the case of floating point computations. This method has been applied both on segmented anatomical structures and on manufactured objects presenting corners and edges. The experimental results show that the method allows to produce a polygonal boundary representation which is guaranteed to be a 2-manifold. This representation is successfully transformed into a triangular quality mesh which meets all topological and geometrical requirements of applications such as augmented reality or simulation.

About this research paper

What this paper is about

We present a discrete approach for boundary extraction from 3D image data. The proposed technique is based on the duality between the Voronoi graph computed accross the digital boundary and the Delaunay triangulation. The originality of the approach is that algorithms perform only integer arithmetic and the method does not suffer from standard round problems and numerical instabilities in the case of floating point computations. This method has been applied both on segmented anatomical structures and on manufactured objects presenting corners and edges. The experimental results show that the method allows to produce a polygonal boundary representation which is guaranteed to be a 2-manifold. This representation is successfully transformed into a triangular quality mesh which meets all topological and geometrical requirements of applications such as augmented reality or simulation.

Why it matters

OpenAlex reports 3 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

We present a discrete approach for boundary extraction from 3D image data. The proposed technique is based on the duality between the Voronoi graph computed accross the digital boundary and the Delaunay triangulation. The originality of the approach is that algorithms perform only integer arithmetic and the method does not suffer from standard round problems and numerical instabilities in the case of floating point computations. This method has been applied both on segmented anatomical structures and on manufactured objects presenting corners and edges. The experimental results show that the method allows to produce a polygonal boundary representation which is guaranteed to be a 2-manifold. This representation is successfully transformed into a triangular quality mesh which meets all topological and geometrical requirements of applications such as augmented reality or simulation.

Key concepts: Delaunay triangulation, Voronoi diagram, Dual graph, Boundary (topology), Boundary representation, Computer science, Bowyer–Watson algorithm, Triangle mesh

Related papers

Back to paper searchBrowse research topicsOriginal source
Discrete Delaunay: Boundary extraction from voxel objects — Research Paper | ScholarLens