20.500.12556/DKUM-97741 Uporaba analize slik isker in umetne inteligence za napovedovanje vsebnosti ogljika v jeklih magistrsko delo Using spark image analysis and artificial intelligence to predict the carbon content of steels V magistrskem delu obravnavamo uporabo slikovne analize isker in metod umetne inteligence za napovedovanje vsebnosti ogljika v jeklih. Na podlagi značilnosti isker, kot so barva, oblika in tekstura, razvijemo napovedne modele z algoritmi strojnega učenja. Za napovedovanje ogljika v jeklih uporabimo tudi konvolucijske nevronske mreže. Rezultati kažejo, da je mogoče z visoko natančnostjo oceniti vsebnost ogljika, kar omogoča hitrejšo in cenejšo analizo v primerjavi s klasičnimi laboratorijskimi metodami. In this master's thesis, we explore the use of spark image analysis and artificial intelligence methods to predict the carbon content in steels. Based on spark characteristics such as color, shape, and texture, we develop predictive models using machine learning algorithms. To estimate carbon content, we also employ convolutional neural networks. The results demonstrate that carbon content can be predicted with high accuracy, offering a faster and more cost-effective alternative to traditional laboratory methods. analiza isker umetna inteligenca lastnosti jekel brušenje napovedni modeli spark analysis artificial intelligence steel properties grinding predictive models true true false [R. Knupleš] Slovenski jezik Angleški jezik Magistrsko delo/naloga 2026-04-09 10:07:16 2026-06-02 14:35:35 2026-06-06 03:06:54 0000-00-00 00:00:00 2026 0 Maribor 0 Maribor 1 spletni vir (1 datoteka PDF (XII, 70 f.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.8/.9:[546.26.062:669.14](043.2) 280633603 MAG_Knuples_Rok_2026.pdf MAG_Knuples_Rok_2026.pdf 1 CA2F4FB69BB1BDB2CF04230AE668A48A 54cc9008da7f5b93e58e0ee20fd71374b81b660db9785972541e5fb6b69dc06b 3f62bb7b-5dcb-11f1-956d-00155d580e34 https://dk.um.si/Dokument.php?lang=slv&id=206622 Fakulteta za strojništvo 0 0 0