Compressive electron backscatter diffraction imaging

• DOI: 10.1111/jmi.13379
• Publication Date: 2025-01-11T10:10:44Z
• Authors: Zoë Broad; Alex W. Robinson; Jack Wells; Daniel Nicholls; Amirafshar Moshta...; Angus I. Kirkland; Nigel D. Browning
• Keywords: Inpainting; Robustness

Abstract Electron backscatter diffraction (EBSD) has developed over the last few decades into a valuable crystallographic characterisation method for wide range of sample types. Despite these advances, issues such as complexity preparation, relatively slow acquisition, and damage in beam‐sensitive samples, still limit quantity quality interpretable data that can be obtained. To mitigate issues, here we propose based on subsampling probe positions subsequent reconstruction an incomplete set. The missing locations (or pixels image) are recovered via inpainting process using dictionary‐learning called beta‐process factor analysis (BPFA). investigate robustness both our Hough‐based indexing, simulate subsampled noisy EBSD sets from real fully sampled Ni‐superalloy set different ratios Gaussian Poisson noise models. We find zero solution pixel detection (inpainting un‐indexed pixels) enables higher‐quality reconstructions to Numerical tests confirm high‐quality band contrast inverse pole figure maps only 10% positions, with potential reduce this 5% if needed. These results show application EBSD, allowing faster extending use technique beam sensitive materials.