Datum održavanja: srijeda, 18.4.2018. u 10:15 sati, prostorija O-358
Predavač: dr. sc. Matej Babič, Slovenija
Naziv predavanja: A novel approach for pattern recognition by using graph theory

Sažetak:

In this work, is presented a novel approach for pattern recognition by using graph theory and its application in mechanical engineering. One of the most important characteristics of the microstructure of a material is the grain size. The analysis of structure is based on statistical methods and on the principle: from the general image of the structure to the particular (individual grains) by averaging. With computer methods, material structure is characterised by analysing each grain. When using computer software tools it is also possible to analyse the area occupied by a certain structure directly, without the need for approximate calculations, as well as automatically create a report file. The aim of this paper is to present a new methodology; a novel approach for pattern recognition by using graph theory and its application in the analysis of grain size distribution values. These values were obtained from SEM images of robot laser-hardened specimens using estimating methods to calculate the fractal dimension of 3D objects and calculating the density of 3D visibility graphs.

Matej Babič received his Ph. D. degree in Computer Science in 2014 from the Faculty of Electrical Engineering and Computer Science of the University of Maribor. Slovenia. He studied Mathematics at the Faculty of Education in Maribor. He concluded postdoc at Jožef Stefan Institute, Ljubljana, Slovenia. His research interest is in fractal geometry, graph theory, intelligent systems, hybrid machine learning and topography of materials after hardening.