2005•Journal of Computing and Information Science in EngineeringRequires access

Contouring Medial Surface of Thin-Plate Structures Using Local Marching Cubes

Tomoyuki Fujimori, Hiromasa Suzuki, Yohei Kobayashi, Kiwamu Kase

Open publisher page 7 citations

Abstract

This paper describes a new algorithm for contouring a medial surface from CT (computed tomography) data of a thin-plate structure. Thin-plate structures are common in mechanical structures, such as car body shells. When designing thin-plate structures in CAD (computer-aided design) and CAE (computer-aided engineering) systems, their shapes are usually represented as surface models associated with their thickness values. In this research, we are aiming at extracting medial surface models of thin-plate structures from their CT data for use in CAD and CAE systems. Commonly used isosurfacing methods, such as marching cubes, are not applicable to contour the medial surface. Therefore, we first extract medial cells (cubes comprising eight neighboring voxels) from the CT data using a skeletonization method to apply the marching cubes algorithm for extracting the medial surface. It is not, however, guaranteed that the marching cubes algorithm can contour those medial cells (in short, not “marching cubeable”). In this study, therefore we developed cell operations that correct topological connectivity to guarantee such marching cubeability. We then use this method to assign virtual signs to the voxels to apply the marching cubes algorithm to generate triangular meshes of a medial surface and map the thicknesses of thin-plate structures to the triangle meshes as textures. A prototype system was developed to verify some experimental results.

About this research paper

What this paper is about

This paper describes a new algorithm for contouring a medial surface from CT (computed tomography) data of a thin-plate structure. Thin-plate structures are common in mechanical structures, such as car body shells. When designing thin-plate structures in CAD (computer-aided design) and CAE (computer-aided engineering) systems, their shapes are usually represented as surface models associated with their thickness values. In this research, we are aiming at extracting medial surface models of thin-plate structures from their CT data for use in CAD and CAE systems. Commonly used isosurfacing methods, such as marching cubes, are not applicable to contour the medial surface. Therefore, we first extract medial cells (cubes comprising eight neighboring voxels) from the CT data using a skeletonization method to apply the marching cubes algorithm for extracting the medial surface. It is not, however, guaranteed that the marching cubes algorithm can contour those medial cells (in short, not “marching cubeable”). In this study, therefore we developed cell operations that correct topological connectivity to guarantee such marching cubeability. We then use this method to assign virtual signs to the voxels to apply the marching cubes algorithm to generate triangular meshes of a medial surface and map the thicknesses of thin-plate structures to the triangle meshes as textures. A prototype system was developed to verify some experimental results.

Why it matters

OpenAlex reports 7 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

This paper describes a new algorithm for contouring a medial surface from CT (computed tomography) data of a thin-plate structure. Thin-plate structures are common in mechanical structures, such as car body shells. When designing thin-plate structures in CAD (computer-aided design) and CAE (computer-aided engineering) systems, their shapes are usually represented as surface models associated with their thickness values. In this research, we are aiming at extracting medial surface models of thin-plate structures from their CT data for use in CAD and CAE systems. Commonly used isosurfacing methods, such as marching cubes, are not applicable to contour the medial surface. Therefore, we first extract medial cells (cubes comprising eight neighboring voxels) from the CT data using a skeletonization method to apply the marching cubes algorithm for extracting the medial surface. It is not, however, guaranteed that the marching cubes algorithm can contour those medial cells (in short, not “marching cubeable”). In this study, therefore we developed cell operations that correct topological connectivity to guarantee such marching cubeability. We then use this method to assign virtual signs to the voxels to apply the marching cubes algorithm to generate triangular meshes of a medial surface and map the thicknesses of thin-plate structures to the triangle meshes as textures. A prototype system was developed to verify some experimental results.

Key concepts: Marching cubes, Contouring, Skeletonization, Fast marching method, Isosurface, Polygon mesh, Surface (topology), Voxel

Related papers

Back to paper searchBrowse research topicsOriginal source
Contouring Medial Surface of Thin-Plate Structures Using Local Marching Cubes — Research Paper | ScholarLens