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 digitisation. To achieve our goals, we rely on our specific research, development and technology competencies, which are the basis of our commitment to excellence in all areas. With our open culture of innovation and our motivated, international teams, we are working to position AIT as Austria's leading research institution at the highest international level and to make a positive contribution to the economy and society.

Our Center for Vision, Automation & Control located in Vienna invites applications for a master's thesis position. The Center for Vision, Automation & Control exploits the opportunities provided by automation and digitisation to initiate and advance innovation for industry, primarily in Austria and Europe. Our main goal is to take away the monotonous, heavy, difficult and dangerous aspects of people’s work through innovation. We help our partners by making technical systems more robust, reliable, flexible and easy to use. At the same time we increase the resource efficiency of industrial processes by reducing waste, product failures and emissions. Our work helps improve product quality and secure Austria’s long-term position as a prime location for industry.

Our Competence Unit High-Performance Vision Systems has been active in research for industrial inspection and quality assurance systems for over 20 years, which creates a stimulating and supportive framework for the students working with us - Therefore, students will have the unique opportunity to work on an interesting real-world problem that arose in ongoing industrial collaborations and international research projects.

• The inline computational imaging (ICI) technique of the Competence Unit High-Performance Vision Systems at AIT is designed for 2D and 3D quality inspection of industrial parts (e.g., parts of cars, PCB boards, complex plastic shapes). To acquire the 3D shape of an object, a camera is moved in reference to the object to capture multi-view images. Additionally, an object is illuminated from different angles to gain information about the surface reflectance behaviour and the local surface orientation, down to a micrometre level. We combine data to generate a highly precise 3D point cloud in a high-speed fashion for finding small surface defects such as scratches or cracks.
• Precise registration with a reference scan or model of the inspected object is pivotal for industrial quality inspection and the control of production processes in this context, since registration will allow for both determining the accuracy of a specific inspection method and assessing a manufactured product’s dimensions and its defect areas in a quantitative fashion.
• The objective of this project is to evaluate state-of-the-art point cloud matching algorithms for 3D shapes and incorporate innovative matching algorithms into a novel point cloud matching and registration tool. The master thesis project will further the research in point cloud registration with a reference object for industrial use case of complex 3D shapes, such as products manufactured by die casting or injection moulding.

• For this master’s thesis project, the following specific aims and related tasks are proposed, including their expected outcomes:
  • (1) Literature research of the state of the art in point cloud registration methods
  • (2) Create a dataset and set up an evaluation framework
  • (3) Evaluation of the point cloud matching algorithms using the set-up framework
  • (4) Create a point cloud matching and registration tool
  • (5) Improving the state-of-the-art. If improvements for current algorithms can be identified during the projects, adapt existing methods.
• This project entails practical programming work and the analysis of state-of-the-art algorithms. Furthermore, it entails hands-on lab work, such as the acquisition of point clouds with our current lab setups. Finally, the identified algorithms will be tested and evaluated against fitting datasets using evaluation metrics.
• We expect that the evaluation of state-of-the-art point cloud alignment methods for 3D point clouds will benefit the insights into the strengths and weaknesses of modern methods for point cloud matching and registration. Developing a novel 3D point cloud fusion based on the findings will improve the quality of current error detection methods for industrial objects with complex 3D shapes.
• You will get familiar with state-of-the-art image processing methods that are at the forefront of application-driven research in the field.