Degree of local symmetry for geometry-aware selective part visualisation on CT volumes. Issue 5 (3rd September 2022)

• Record Type: Journal Article
• Authors: Yasunami, Nobumichi; Yatagawa, Tatsuya; Ohtake, Yutaka; Suzuki, Hiromasa
• Is Part Of: Nondestructive testing and evaluation. Volume 37:Issue 5(2022)
• Page Start: 666
• Page End: 678
• Publication Date: 2022-09-03
• Subjects: X-ray CT volume -- parts visualization -- volume segmentation -- geometric features; Non-destructive testing -- Periodicals; 620.112705
• DOI: 10.1080/10589759.2022.2090556
• Languages: English
• ISSNs: 1058-9759

Abstract

ABSTRACT: Visualising the complex assembly in a large CT volume is a difficult task due to the existence of components made of materials with similar X-ray attenuation factors. To resolve the difficulty, we propose a degree of local symmetry (DLS), a new voxel attribute to visualise specific parts based on their local geometric features. The DLS represents a dimensionality of the manifold around a voxel ( e.g . the DLS for a local surface will be 2). For a given CT volume, we first extract an isosurface of the target assembly imaged by an X-ray CT device and compute the medial axes of the isosurface. We then compute the DLS values for voxels on the medial axes. Finally, the DLS values are propagated from the voxels on the medial axes to the entire voxels of the input volume. We demonstrate that the DLS works well to selectively visualise the parts in various assemblies, including a simple artificial assembly of several primitive shapes and more complicated vehicle models.