Interactive volumetric segmentation for textile micro‐tomography data using wavelets and nonlocal means

Abstract

This work addresses segmentation of volumetric images of woven carbon fiber textiles from micro‐tomography data. We propose a semi‐supervised algorithm to classify carbon fibers that requires sparse input as opposed to completely labeled images. The main contributions are: (a) design of effective discriminative classifiers, for three‐dimensional textile samples, trained on wavelet features for segmentation; (b) coupling of previous step with nonlocal means as simple, efficient alternative to the Potts model; and (c) demonstration of reuse of classifier to diverse samples containing similar content. We evaluate our work by curating test sets of voxels in the absence of a complete ground truth mask. The algorithm obtains an average 0.95 F1 score on test sets and average F1 score of 0.93 on new samples. We conclude with discussion of failure cases and propose future directions toward analysis of spatiotemporal high‐resolution micro‐tomography images.

Flow Mechanics in Ablative Thermal Protection Systems

“…An ongoing challenge in µ-CT images of strongly anisotropic materials, such as weaves or fiber composites, is segmentation of the phases and determination of fiber and tow orientation. While these remain active areas of research, several methods to determine fiber orientation voxelwise from µ-CT have been developed (Krause et al 2010, MacNeil et al 2019, Semeraro et al 2020. Also noteworthy are recent advances in the quantification of uncertainties on effective properties calculated from 3D images due to segmentation errors (Krygier et al 2021).…”

Quantifying the unknown impact of segmentation uncertainty on image-based simulations

“…Image-based simulation has become a crucial part of modern engineering analysis workflows, alongside integrated computational materials engineering (ICME) 1 and digital twin [2][3][4][5] efforts. It has also been adopted by many disciplines, including medical imaging for patient treatment [6][7][8][9][10] , improving the manufacturing process of batteries [11][12][13][14] , composite material development [15][16][17][18] , biological physics 19 , neuroscience 20;21 , and geomechanics 22 Image-based simulation traditionally involves three steps, illustrated in Figs. 2a-2d.…”

Quantifying the Unknown: Impact of Segmentation Uncertainty on Image-based Simulations.

“…Image-based simulation has become a crucial part of modern engineering analysis workflows, alongside integrated computational materials engineering (ICME) 1 and digital twin [2][3][4][5] efforts. It has also been adopted by many disciplines, including medical imaging for patient treatment [6][7][8][9][10] , improving the manufacturing process of batteries [11][12][13][14] , composite material development [15][16][17][18] , biological physics 19 , neuroscience 20,21 , and geomechanics 22 .…”