Smart Align
The Oxford invention enables image-processing, enhancement and quantification in dark and bright-field scanning transmission electron microscopy (STEM) and conventional TEM. It uses a weighted learning filter to guide the rigid-registration stage. Built-in knowledge describes the scanning nature of the serial acquisition and reduces artefact introduction. As a result, this technology delivers superior registration results compared with current methods, while very little human intervention is required.
Motion sickness
The ‘Smart Align’ software delivers superior image processing, enhancement and quantification in dark- and bright-field scanning transmission electron microscopy (STEM), conventional TEM registration and scanning tunnelling microscopy (STM). It requires very little human intervention. The software uses a weighted learning filter to guide the rigid-registration stage. Built-in knowledge describes the scanning nature of the serial acquisition and reduces artefact introduction. A mature demo version is available for assessment.
Automated registration
STEM data is recorded serially and acquisition times can be tens of seconds long. At these acquisition times, stage/sample drift and low-frequency distortions can perturb the image locally. Often the first step in any quantitative interpretation of STEM data is to correct for these drifts and distortions using so-called rigid and non-rigid registrations respectively.
The Oxford invention is an improved automated method of performing this registration step, customised for the challenges unique to STEM data. The improvements developed to address the challenges of registering images which contain a large proportion of crystalline material and/or local features of interest such as dislocations or edges.
Demonstration available
‘Smart Align’ uses a new learning model for rigid registration of images dominated by periodic (crystallographic) features and a ‘row-locking’ mode for artefact free non-rigid registration of serially acquired data. It includes novel options designed with EM image processing in mind. The software has
been rigorously tested and developed to a mature demonstration version.
Superior results
The Oxford invention enables processing and enhancement of microscopy images, delivering superior results and quantification compared with current methods. Through built-in knowledge and a weighted learning filter, this is achieved with very little human input. The technology has been specifically developed for the analysis of dark-field scanning transmission electron microscopy images but can be used with bright-field data, conventional STEM and in other image analysis applications.
We are partnered with HREM who maintain and distribute this software. Please contact HREM for further information:
about this technology
