The new CmarT project aims to advance tomographic data acquisition and processing at leading synchrotron facilities such as PETRA III and ESRF. By combining advanced physical modeling with cutting-edge reconstruction algorithms for multi-scale and hierarchical imaging, it will dramatically increase the output data quality. Efficient implementation and integration into the experiment workflows will transform the way scientists conduct complex 3D imaging experiments.

X-ray microscopy and tomography at synchrotron facilities enable high-resolution imaging of specimens ranging from single cells to bulk materials. However, fully exploiting these methods is challenging due to complex image acquisition, reconstruction, and analysis, especially for large samples. Current methods require extensive processing to extract relevant features, making experiments time-consuming and difficult for non-experts. The CmarT project aims to develop local adaptive tomography (LAT), an approach that combines adaptive data acquisition with advanced reconstruction algorithms and allows users to efficiently scan large, complex samples and at the same time capture important fine details. The resulting infrastructure will be integrated into experimental control and enable fast, interactive experiments with high-quality output data.

Organizational unit

Laboratorium für Applikationen der Synchrotronstrahlung (LAS)

Job description

Main Tasks:

• Implement and optimize novel phase retrieval, artifact-reduction and motion compensation algorithms and integrate them into 3D reconstruction pipelines
• Develop metrics for real-time feedback on scan quality, including 2D and 3D resolution estimates.
• Integrate the developed pipelines into the beamline control system

Starting date

01.03.2026

Personal qualification

PhD in computer science, mathematics, or related disciplines or master degree with 2+ years of experience with X-ray imaging and image processing. Required Skills: Good proficiency in programming, particularly with C and Python; work experience with image-processing; fluency with Linux; experience with 3D X-ray imaging, machine learning, and parallel programming with OpenCL/CUDA is a plus.