Abstract
The massive amount of diffraction images collected in a raster scan of Laue microdiffraction calls for a fast treatment with little if any human intervention. The conventional method that has to index diffraction patterns one-by-one is laborious and can hardly give real-time feedback. In this work, a data mining protocol based on unsupervised machine learning algorithm was proposed to have a fast segmentation of the scanning grid from the diffraction patterns without indexation. The sole parameter that had to be set was the so-called “distance threshold” that determined the number of segments. A statistics-oriented criterion was proposed to set the “distance threshold”. The protocol was applied to the scanning images of a fatigued polycrystalline sample and identified several regions that deserved further study with, for instance, differential aperture X-ray microscopy. The proposed data mining protocol is promising to help economize the limited beamtime.
Highlights
Laue diffraction, that may occur when a polychromatic X-ray beam illuminated a crystal, was first discovered in 1912, and has unveiled both the electromagnetic nature ofX-ray and the periodic ordering of atoms in crystal [1]
We found that negative matrix factorization (NMF) costed much more computation time than principal component analysis (PCA)
An indexation-free treatment of raster scanning Laue microdiffraction patterns was proposed to have a fast segmentation of the scanning grids based on their underlying diffraction patterns, whereas the conventional treatment had to index diffraction patterns one-by-one and could hardly give on-the-fly feedback
Summary
That may occur when a polychromatic X-ray beam illuminated a crystal, was first discovered in 1912, and has unveiled both the electromagnetic nature ofX-ray and the periodic ordering of atoms in crystal [1]. That may occur when a polychromatic X-ray beam illuminated a crystal, was first discovered in 1912, and has unveiled both the electromagnetic nature of. Thanks to the polychromaticity of the employed X-ray, multiple diffraction peaks can be recorded in a single exposure without any rotation, thereby excluding the ambiguity of the illuminated volume [2]. Compared to the electron backscatter diffraction (EBSD) technique, Laue microdiffraction technique functions by raster scanning the sample to generate the lattice orientation and distortion maps from the one-by-one analysis of the diffraction pattern emanating from each scanned spot [4,5,6]. It is generally accepted that EBSD has an edge on finer spatial resolution of nanoscale, whilst Laue microdiffraction can have a much better accuracy on the lattice orientation and distortion with an attainable order of 10−9 [9]
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call