IPE 02-21 Master Thesis or Internship: Compensation of laminograpic artifacts with regularized iterative reconstruction
Institute for Data Processing and Electronics (IPE)
X-ray Computed Tomography (CT) is a widely used three-dimensional imaging method which employs penetrating capabilities of photons and dedicated mathematical algorithms to visualize and analyze the internal structures. Laminography is a class of tomographic imaging method which is designed for flat and laterally extended samples/objects in which the rotation axis is tilted relative to the beam direction, such as PCBs, composite panels, flat materials and even paintings. Laminography provides more uniform exposure of planar samples and good reconstruction of features located in the plane perpendicular to rotation axis. However, conventional tomographic methods are not suitable for laminography and, if not adapted, cause severe artifacts in the reconstructed images, shown in the image. These artifacts prevent proper image segmentation and quantitative analysis and present a significant bottleneck for both scientific and industrial applications.
A viable way to reduce the laminographic artifacts is by means of regularized iterative reconstruction. Iterative methods formulate reconstruction as an optimization problem consisting of a data fidelity term and a regularizer. The latter term encodes prior knowledge about the image and helps to compensate for insufficient data – the main cause of the laminographic artifacts. There is no unique regularizer which performs well for all tomographic problems and it has to be carefully selected depending on the underlying tomographic data properties. In this exciting project you will work on the edge between mathematics, computer science, engineering and physics. You will explore state-of-the-art image reconstruction techniques in a quest for a robust technique to improve laminographic image quality. You will use image reconstruction frameworks to tailor regularization techniques and test your findings on data acquired on various synchrotron light sources. Your work will also benefit from our highly interdisciplinary team and exposure to synchrotron imaging community.
The project is offered as both internship or Master thesis. In case of internship, you will assist researchers in multiple projects and will get exposure to different types of data and regularization methods. A specific project will be assigned for students seeking to write Master thesis. They will work independently and will be responsible for analyzing the data, suggesting appropriate regularization techniques, and quantifying obtained results. The details will be discussed during interview.
as soon as possible
Internship is suited for students majoring in physics, mathematics, or computer science. The applicants are expected to be verse in image processing techniques using MATLAB or Python. The prior exposure to computed tomography or/and mathematical optimization techniques is a plus.
Expected duration: 6-months
Suren Chilingaryan firstname.lastname@example.org
Phone: +49 721 / 608 26579
Evelina Ametova email@example.com
Please apply online using the button below for this vacancy number IPE 02-21.
Ausschreibungsnummer: IPE 02-21
Recognized severely disabled persons will be preferred if they are equally qualified.
Personnel Support is provided by:
Phone: +49 721 608-25184,
Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany