Master thesis: Real-time image analysis for the identification of defects i

The research institute Swerim provides applied research within mining engineering, process metallurgy and materials and manufacturing engineering, mainly for the mining, steel and metals industry. Swerim has 200 co-workers in two locations in Sweden - Luleå and Stockholm.

OverviewSteel bars are rolled into wires at high speeds and using multiple subsequent rolling operations, as the wire gets thinner. Inspecting the wire surface during production tells much about the wire quality, but is difficult to do this manually due to the high speed of the wire in production. Using a machine learning-based approach to fast classify camera images of the wire surface may help in tuning or otherwise improving the wire rolling operations.

A modern deep learning-based image recognition algorithm may be able to solve this problem. The project aims at applying and adapting exiting algorithms, using available open-source packages to classify in production wire surface.


Figure 1 Defect on hot rod (first and second from left) and hot wire (third and fourth from left) with to different filters applied.
GoalIdentify and design suitable methods for fast analysis of data sampled online in order to detect flaws and/or defects on material surface in real-time. Data is collected with a camera system and saved in avi-format. The defects have different characteristics and in order to make them all visible a number of filters might be needed to apply during analysis. Data is collected with a LabVIEW-based system, but analysis could be done in a preferable environment (i.e., Matlab or Python).

Further information
Swerim rewards the student with 50 000 SEK for an approved master thesis (30hp).  

Application
Apply by using the application function below. You will receive a confirmation that Swerim has received your application. Please note that we fill the thesis as soon as we find a suitable applicant, which means we can fill the position before the deadline.

The work is planned to start January 2023 at Swerim in Stockholm.

Contact
For further information about project, please contact: 
Peter Lundin, peter.lundin@swerim.se