Skip to content
Vivian edited this page May 6, 2026 · 6 revisions

page under construction

Welcome to the mtex-trueEbsd wiki!

TrueEBSD is an MTEX add-on for spatial alignment and distortion correction of EBSD maps and/or images.

Who should use TrueEBSD?

TrueEBSD could help you if:

  • you have a set of images and/or data maps of the same physical features, and you want to align them pixel-wise
  • you expect a specific distortion to have a certain shape, and you want the fit to be physically realistic

TrueEBSD is particularly useful if:

  • you have a large amount of data to align, where manual selection of control point pairs would be extremely time-consuming
  • you need to align data maps (e.g. EDS or EBSD data) as well as images
  • you have compounding distortion types in your images (e.g. drift + tilt distortions in an EBSD map)
  • you want to analyse your data in MTEX after processing with TrueEBSD

Examples of where TrueEBSD has been useful: [TODO - add images]

  1. Combining an SEM-SE image with EBSD to automate crystallographic slip plane indexing from surface traces [link]
  2. Combining an SEM-BSE image with EBSD to determine preferred grain boundary void distributions in a copper polycrystal [link]
  3. Combining SEM-BSE images with EBSD to improve phase segmentation and boundary analysis in WC-Co [link]
  4. Correcting EBSD map tilt and drift distortions (relative to a flat SEM image) to combine multiple EBSD maps along a 2mm-long crack in WC-Co, and determine the preferred fatigue crack growth planes [link]
  5. Combining three EBSD maps acquired from different EBSD systems, to measure the absolute orientation uncertainty in EBSD. [link]
  6. Correcting EBSD map tilt and drift distortions relative to a flat SEM image, to enable seamless stitching of large-area tiled EBSD maps [link]
  7. Combining polarised light optical micrographs with EBSD to determine preferred grain and grain boundary locations for Zr hydride growth in Zircaloy-4 [link]
  8. Combining two EBSD maps, before and after SEM in-situ tensile testing, to determine slip activity from grain reorientation axes in an in-situ deformed Zr polycrystal [link]

TrueEBSD might not help if**:

  • your images have flat contrast or only very sparse features, e.g. you are imaging a single crystal with minimal topography
  • your images do not contain any recognisable shared features, e.g. if one of your images is a SEM image of a polycrystal with grain orientation contrast, and the other is an indentation hardness map with step size larger than the average grain size
  • your images are not of the same physical surface, e.g. serial sections of an image volume**
  • the distortion field in your image is not well-described with a mathematical distortion model, e.g. projective tilt distortions of a rough surface**
  • the distortion field in your image is discontinuous, e.g. 'jumps' in an EBSD map due to electron beam charge/discharge artefacts**

**TrueEBSD could be still worth a try, and some of these are in the development pipeline. Contact me for further support!

TrueEBSD does not:

  • interpolate or modify values of non-image data (such as EBSD orientation or EDS element maps)
  • run on older MTEX versions < 6

Getting started

To start using TrueEBSD, you will need:

Experimental data:

  • A distorted image, e.g. an EBSD map
  • A reference ground truth image, e.g. a SEM image
  • Knowledge of the distortion types between image pairs, e.g. drift (aka rolling-shutter distortion), affine, projective
  • [optional] Intermediate image(s), which may be used to separate out compound distortions into individual types.

By default, cross-correlation of the images' edge transforms are used to compute local shifts in each part of the image, which are fitted to the specified distortion model.

System requirements

  • MTEX 6.1+
  • 64-bit Windows or Linux [or Parallel Computing Toolbox]
  • MATLAB 2024a+
  • Ghostscript (free software for PDF export https://ghostscript.com/)
  • MATLAB toolboxes:
    • Curve Fitting Toolbox
    • Image Processing Toolbox
    • Statistics and Machine Learning Toolbox

Limitations

Example data

Resources

Contact