Vacancy No. IPE 13-17
IPE 13-17 / Master Thesis in Computer Science: GPU-assisted tracking of single bubbles in multiphase flows
Multiphase flows in fluidized beds are complex and difficult to observe because of their dynamics. But exact information on gas distribution and fluid transport is essential for many chemical processes, such as fluid catalytic cracking and polymerization, drying, combustion, granulation, purification or coating. To study such flows, the ultrafast X-ray tomography system ROFEX has been developed at TU Dresden. With frame rates of up to 10,000 fps the fast processes even in opaque media can be monitored in 3D. At the downside the high frame rates result in large data streams that need to be analyzed online in order to track single bubbles and to control the scanner position according to movement of the tracked structures along the height of the column.
The goal of the Master thesis is to develop and characterize a set of imaging algorithms required to track a single bubble within a bubble swarm in the column. A sequence of image processing steps has to be performed online: The images need to be reconstructed from the raw data acquired by the ROFEX detector. The bubble has to be identified and the control decision how to move the detector has to be made. The challenge is to guarantee 80 control decisions per second to be able to achieve real-time control. For scenarios with turbulences even a higher decision rate may be required.
Good background in computer vision, strong knowledge of C/C++ programming. Prior experience in parallel programming using CUDA or OpenCL is a plus.
Institute for Data Processing and Electronics (IPE)
Dr. Suren Chilingaryan <suren.chilingaryan@kit edu> Phone: +49 721 / 608 26579
Dr. Andreas Kopmann <andreas.kopmann@kit edu> Phone: +49 721 / 608 24910
Please apply online using the button below for this vacancy number IPE 13-17.
Personnel support is provided by
phone: +49 721 608-25184,
Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany
If qualified, handicapped applicants will be preferred.