Extraction of Curved Fibers From 3d Data

Abstract

A segmentation algorithm is proposed which automatically extracts single fibers from tomographic 3D data of fiber-based materials. As an example, the algorithm is applied to a non-woven material used in the gas diffusion layer of polymer electrolyte membrane fuel cells. This porous material consists of a densely packed system of strongly curved carbon fibers. Our algorithm works as follows. In a first step, we focus on the extraction of skeletons, i.e., center lines of fibers. Due to irregularities like noise or other data artefacts, it is only possible to extract fragments of center lines. Thus, in a second step, we consider a stochastic algorithm to adequately connect these parts of center lines to each other, with the general aim to reconstruct the complete fibers such that the curvature properties of real fibers are reflected correctly. The quality of the segmentation algorithm is validated by applying it to simulated test data.

Statistical Inference for Curved Fibrous Objects in 3D – Based on Multiple Short Observations of Multivariate Autoregressive Processes

“…The stochastic models developed in Gaiselmann et al . (, b, ) are primarily designed for use in materials science, where fibre‐based materials play an important role. The applications of fibre‐based materials include thermal insulation, aircraft and car bodies, and fuel cell technology.…”

Stochastic 3D modeling of non-woven materials with wet-proofing agent

“…The experimental data and the system of fibers extracted from it are in good optical accordance. In [9], it is shown more formally that the algorithm extracts fibers accurately with respect to curvature characteristics. In particular, for a representative test volume of fiber-based material, the algorithm reproduced 91% of connections correctly.…”

“…This algorithm combines tools from image processing and stochastic optimization. It is described in detail in [9] and thus we only mention the basic idea.…”

“…More precisely, the parameters of the non-woven GDL model are fitted to 3D image data of experimental non-woven GDL, where the image data is gained by synchrotron tomography. To adequately fit the model parameters, we extract a system of single fibers from experimental image data of non-woven GDL using an algorithm described in [9]. From the extracted fiber system, important properties like curvature measures or directional distributions of the fibers can be incorporated into the model.…”

Estimating Torsion of Digital Curves Using 3d Image Analysis

“…In Peyrega and Jeulin (2013) the tortuosity is estimated from µCT images of Thermisorel structure (a fiber material applied for acoustic absorption), where special attention is paid to the reconstruction of geodesic paths. We also refer to the paper Gaiselmann et al (2013) where the tortuosity of the fibers of a non-woven material was computed.…”