As Austria's largest research and technology organisation for applied research, we are dedicated to make substantial contributions to solving the major challenges of our time, climate change and digitalisation.

Our Center for Vision, Automation & Control located in Vienna invites applications for a master’s thesis position. The Center exploits automation and digitalisation to initiate and advance innovation for industry, primarily in Austria and Europe.

For this master's thesis project, you will join the Competence Unit " High-Performance Vision Systems (HVS) ”, which has over 20 years of research experience in industrial inspection and quality assurance. At HVS we innovate 2D & 3D vision systems that go beyond the state of the art by developing advanced vision sensor concepts, scalable embedded vision systems, and machine learning methods for industrial high-performance visual quality inspection.

The objective of this master’s thesis project is to evaluate the potential of material aware 3D reconstruction methods for industrial fast inline inspection applications. While many state-of-the-art optical inline 3D scanners, such as laser triangulation sensors, primarily deliver geometric depth information, industrial processes often require deeper object knowledge beyond geometrical dimensions. Color, roughness, reflectance behavior, and many other optically detectable material properties carry essential information for robust defect detection and quality control.

Within our HVS unit we established a top-notch technology for inline computationally imaging color and photometric information alongside 3D depth information. Building on this, the master’s thesis, which will be carried out in cooperation with TU Wien – Research Group Photogrammetry will combine application-oriented research with academic expertise in photogrammetry, remote sensing, and advanced 3D reconstruction methods. The specific thesis’ focus will be adapted depending on the students’ skill and interests, for e.g. towards the integration of material-aware gaussian splatting representation or multispectral imaging.

• You will acquire basic skills in inline optical 3D imaging for process control and automation.
• You will carry out a literature search and review on state-of-the-art optical detection and reconstruction of material properties.
• You will explore the potential of material aware Gaussian Splatting and related approaches or sensing methods like multispectral imaging for fast inline quality control.
• Under the guidance of our experienced researchers, you will have access to our machine vision lab as well as a digital-twin blender model of the lab-setup, to carry out experiments and collect datasets.

• Depending on the direction of the thesis, there is the possibility to expand the existing lab-setup with multispectral sensing capabilities.
• A selected Gaussian Splatting algorithm (e.g. RefGS, 3DGSDR, or GShader) can serve as a starting point for further adaptation towards modelling and reconstructing material properties and can be extended or complemented depending on your individual research direction.