3.3.3.109. NXimage_set_em_kikuchi¶
Status:
base class, extends NXobject
Description:
Measured set of electron backscatter diffraction data, aka Kikuchi pattern. ...
Measured set of electron backscatter diffraction data, aka Kikuchi pattern. Kikuchi pattern are the raw data for computational workflows in the field of orientation (imaging) microscopy. The technique is especially used in materials science and materials engineering.
Based on a fuse of the M. A. Jackson et al. of the DREAM.3D community and the open H5OINA format of Oxford Instruments P. Pinard et al.
EBSD can be used, usually with FIB/SEM microscopes, for three-dimensional orientation microscopy. So-called serial section analyses. For a detailed overview of these techniques see e.g.
With serial-sectioning this involves however always a sequence of measuring, milling. In this regard, each serial section (measuring) and milling is an own NXevent_data_em instance and thus there such a three-dimensional characterization should be stored as a set of two-dimensional data, with as many NXevent_data_em instances as sections were measured.
These measured serial sectioning images need virtually always post-processing to arrive at the aligned and cleaned image stack before a respective digital model of the inspected microstructure can be analyzed. The resulting volume is often termed a so-called (representative) volume element (RVE). Several software packages are available for performing this post-processing. For now we do not consider metadata of these post-processing steps as a part of this base class because the connection between the large variety of such post-processing steps and the measured electron microscopy data is usually very small.
If we envision a (knowledge) graph for EBSD it consists of individual sub-graphs which convey information about the specimen preparation, the measurement of the specimen in the electron microscope, the indexing of the collected Kikuchi pattern stack, eventual post-processing of the indexed orientation image via similarity grouping algorithms to yield (grains, texture). Conceptually these post-processing steps are most frequently serving the idea to reconstruct quantitatively so-called microstructural features (grains, phases, interfaces). Materials scientists use these features according to the multi-scale materials modeling paradigm to infer material properties. They do so by quantifying correlations between the spatial arrangement of the features, their individual properties, and (macroscopic) properties of materials.
Symbols:
n_sc: Number of scanned points. Scan point may have none, one, or more pattern.
n_p: Number of diffraction pattern.
n_y: Number of pixel per Kikuchi pattern in the slow direction.
n_x: Number of pixel per Kikuchi pattern in the fast direction.
Structure:
PROCESS: (optional) NXprocess
Details how Kikuchi pattern were processed from the detector readings. ...
Details how Kikuchi pattern were processed from the detector readings. Scientists interested in EBSD should inspect the respective NXem_ebsd application definition which can be used as a partner application definition to detail substantially more details to this processing.
stack: (optional) NXdata
Collected Kikuchi pattern as an image stack. As raw and closest to the ...
Collected Kikuchi pattern as an image stack. As raw and closest to the first retrievable measured data as possible, i.e. do not use this container to store already averaged, filtered or whatever post-processed pattern unless these are generated unmodifiably by the instrument given the way how the instrument and control software was configured for your microscope session.
scan_point_identifier: (optional) NX_UINT (Rank: 1, Dimensions: [n_p]) {units=NX_UNITLESS}
Array which resolves the scan point to which each pattern belongs. ...
Array which resolves the scan point to which each pattern belongs. Scan points are evaluated in sequence starting from scan point zero until scan point n_sc - 1. Evaluating the cumulated of this array decodes which pattern in intensity belong to which scan point. In an example we may assume we collected three scan points. For the first we measure one pattern, for the second we measure three pattern, for the last we measure no pattern. The values of scan_point_identifier will be 0, 1, 1, 1, as we have measured four pattern in total.
In most cases usually one pattern is averaged by the detector for some amount of time and then reported as one pattern. Use compressed arrays allows to store the scan_point_identifier efficiently.
data_counts: (optional) NX_NUMBER (Rank: 3, Dimensions: [n_p, n_y, n_x]) {units=NX_UNITLESS} ⤆
Signal intensity. For so-called three-dimensional or serial sectioning ...
Signal intensity. For so-called three-dimensional or serial sectioning EBSD it is necessary to follow a sequence of specimen surface preparation and data collection. In this case users should collect the data for each serial sectioning step in an own instance of NXimage_set_em_kikuchi. All eventual post-processing of these measured data should be documented via NXebsd, resulting microstructure representations should be stored as NXms.
@long_name: (optional) NX_CHAR ⤆
Kikuchi pattern intensity
pattern_identifier: (optional) NX_UINT (Rank: 1, Dimensions: [n_p]) {units=NX_UNITLESS}
Pattern are enumerated starting from 0 to n_p - 1.
@long_name: (optional) NX_CHAR
Kikuchi pattern identifier
axis_y: (optional) NX_NUMBER (Rank: 1, Dimensions: [n_y]) {units=NX_LENGTH}
axis_x: (optional) NX_NUMBER (Rank: 1, Dimensions: [n_x]) {units=NX_LENGTH}
Hypertext Anchors¶
List of hypertext anchors for all groups, fields, attributes, and links defined in this class.
