Department Microstructure and Residual Stress Analysis

AI & Machine learning

The main objective of this research activity is the improvement of the imaging technique X-ray tomoscopy (time-resolved tomography) developed at Helmholtz-Zentrum Berlin acting as a customer. The project addresses the improvement of time resolution, image quality and quantitative output by advanced image analysis, as well as the development and implementation of deep learning algorithms to boost image evaluation speeds as provided by the expertise of the German Aerospace Center. We raise the challenge to upgrade the current tomoscopy method by dedicated image analysis in a complementary cooperation between the partners in a cross-centre and cross-field approach to strengthen the possibilities of the method. This approach extends tomoscopy to have a wide impact in other research areas such as geology, biology, medicine, etc.