Data processing methods and data acquisition for samples larger than the field of view in parallel-beam tomography

Area of Science:

• Synchrotron-based imaging
• Materials science
• X-ray computed tomography

Background:

• Parallel-beam tomography systems at synchrotrons have a limited field of view (FOV).
• Scanning samples larger than the FOV requires complex acquisition schemes like grid or helical scans.
• Existing methods lack easy-to-use, automated software for processing these large datasets.

Purpose of the Study:

• To develop and implement automated software solutions for processing large-scale parallel-beam tomography data.
• To enable high-quality image reconstruction from complex scanning strategies.
• To expand the applicability of synchrotron tomography for oversized samples.

Main Methods:

• Development of a Python package named Algotom.
• Implementation of methods for automated data processing, including image stitching and center-of-rotation (COR) determination.
• Focus on efficient and high-quality reconstruction of data acquired using grid, 360-degree offset COR, and helical scans.

Main Results:

• Algotom provides efficient and automated tools for processing challenging tomography datasets.
• The software facilitates high-quality image reconstruction from various large-sample scanning techniques.
• Demonstrated practicality for synchrotron facilities where manual data processing is infeasible.

Conclusions:

• Algotom significantly enhances the usability and scope of parallel-beam tomography at synchrotrons.
• The package addresses critical software gaps for processing large FOV datasets.
• This work empowers researchers to analyze larger and more complex samples with greater ease and quality.