A modular U-Net for automated segmentation of X-ray tomography images in composite materials

João P C Bertoldo (1, 2) , Etienne Decencière (3, 2) , David Ryckelynck (1, 2) , Henry Proudhon (1, 2)

1 CDM - Centre des Matériaux
2 PSL - Université Paris Sciences et Lettres
3 CMM - Centre de Morphologie Mathématique

Abstract X-ray Computed Tomography (XCT) techniques have evolved to a point that high-resolution data can be acquired so fast that classic segmentation methods are prohibitively cumbersome, demanding automated data pipelines capable of dealing with non-trivial 3D images. Deep learning has demonstrated success in many image processing tasks, including material science applications, showing a promising alternative for a human-free segmentation pipeline. In this paper a modular interpretation of U-Net (Modular U-Net) is proposed and trained to segment 3D tomography images of a three-phased glass fiber-reinforced Polyamide 66. We compare 2D and 3D versions of our model, finding that the former is slightly better than the latter. We observe that human-comparable results can be achievied even with only 10 annotated layers and using a shallow U-Net yields better results than a deeper one. As a consequence, Neural Network (NN) show indeed a promising venue to automate XCT data processing pipelines needing no human, adhoc intervention.

Mots clés: Deep Learning; U-Net; Modular Network Architecture; Semantic Segmentation; 3D X-ray Computed Tomography; Composite Material

Domaines: Mécanique des matériaux [physics.class-ph]

João P C Bertoldo, Etienne Decencière, David Ryckelynck, Henry Proudhon. A modular U-Net for automated segmentation of X-ray tomography images in composite materials. 2021. ⟨hal-03380106v1⟩