Master's thesis: Segmentation of Liquid Foam bubbles in Volumetric X-Ray Computed Tomography Datasets Using Deep Learning

Background

The thesis forms part of the VINNOVA financed project “AI-Tomo: Accelerated materials characterisation by AI and X-ray tomography”. The goal of AI-TOMO is to develop AI algorithms for fast, effective segmentation and quantification of 3D and 4D X-ray tomography data to accelerate materials development. AI-TOMO is a close collaboration between the research providers RISE and Lund University, the synchrotron facility MAX IV, and the companies Billerud and TetraPak. This master thesis will be a collaboration between AI researchers at Lund University and RISE with support from X-ray tomography experts at Lund University.

Liquid foam is extensively used in daily life and in industry for its peculiar solid-liquid behavior: from shaving foam, whipped cream, to large-scale processes such as soil remediation, water treatment and paper recycling. Its well-defined microstructure makes it an ideal model system for studying structural phenomena shared with other disordered materials, such as glasses and biological tissues.

This project focuses on the experimental study of liquid foam structural evolution using X-ray tomography. It involves the segmentation and tracking of individual bubble across consecutive time steps. The current analysis approaches rely on traditional segmentation and tracking methodologies, i.e. Watershed and Digital Image Correlation (DIC) algorithms, which face limitations. The project will develop new tools based on state-of-the-art deep learning architectures, including convolutional and physics-informed neural networks, to overcome these limitations with the segmentation methods. Successful implementation of these techniques will expand the range of accessible time and space resolutions for studying liquid foam structural dynamics.

Goal

To develop deep learning methodologies for segmenting liquid foam bubbles in X-Ray Computed Tomography (XCT) datasets. The goal is to segment each individual foam bubble and track them as they evolve during experimentation inside an X-ray Computed Tomography scanner. The project therefore encompasses both 3D and 4D volumetric segmentation of liquid foam bubbles.

Qualifications

To be able to successfully contribute to this degree project, we believe that you

• Are a master student in computer science, engineering physics or similar,

• Have a background in machine learning, AI or similar with an interest in material science and imaging,

• Have solid oral and written English skills.

Terms

• Recruiting manager: Martin Körling

• Supervisor:
  Foam system: Florian Schott, florian.schott@solid.lth.se
  ML: Smita Chakraborty smita.chakraborty@ri.se, Eoin Walsh, eoin.walsh@solid.lth.se

• Examiner: Prof. Stephen Hall, Division of Solid Mechanics at LTH.

• Location: Primarily at the Division of Solid Mechanics at LTH, Lund University and RISE Lund.

• Application deadline: 15th July 2026.

• Timeline: August/September 2026-January 2027. A detailed schedule will be defined together with the candidate. The expected time is 20 weeks full time (40 hours/week) work, roughly.

• Credits: 30 HP

• Compensation: 30000 SEK upon successful completion of a high-quality thesis.

Contact

For any questions, please contact Smita Chakraborty smita.chakraborty@ri.se, Eoin Walsh eoin.walsh@solid.lth.se, and Florian Schott florian.schott@solid.lth.se.

Scientific supervision will be given by:

Eoin Walsh, Florian Schott (Lund University) and Smita Chakraborty (RISE, Computer Science).

Welcome with your application!

Last day of application: 15th July 2026