Nadzor kakovosti montaže bencinskih katalizatorjev s strojnim vidom

Kokalj, Jakob (avtor), Herakovič, Niko (mentor), Šimic, Marko (komentor)

Magistrsko delo obravnava kontrolo kakovosti montaže s strojnim vidom, kar postaja vse pomembnejši segment pri montaži avtomobilskih katalizatorjev. V nalogi je zaradi posebne vrste katalizatorja obstoječi algoritem kontrole kakovosti nadgrajen tako, da lahko ploščino kosa razdelimo v dve območji in nato vsako območje posebej obravnavamo z drugačnimi kriteriji. V začetku naloge so predstavljeni katalizatorji, osnove kontrole kakovosti in obdelave slik ter opis obstoječega sistema za pregledovanje katalizatorjev. V tem delu je razdelan tudi pregled trga za morebitne rešitve. V drugem delu naloge je opisan potek razvijanja algoritma in delo potrebno za integracijo. Obravnavane so tudi nekatere druge poti raziskovanja z MATLAB-om. V zaključku naloge je prikazano testiranje algoritma in analiza rezultatov.

• Jezik: Slovenski jezik
• Ključne besede: - strojni vid - nadzor kvalitete - segmentacija slik - katalizator - algoritem za iskanje napak
• Vrsta gradiva: Magistrsko delo/naloga
• Organizacija: FS - Fakulteta za strojništvo
• Leto izida: 2017
• PID: 20.500.12556/RUL-97611
• Datum objave v RUL: 28.10.2017

KOKALJ, Jakob, 2017, Nadzor kakovosti montaže bencinskih katalizatorjev s strojnim vidom [na spletu]. Magistrsko delo. [Dostopano 26 marec 2025]. Pridobljeno s: https://repozitorij.uni-lj.si/IzpisGradiva.php?lang=slv&id=97611

Sekundarni jezik

• Jezik: Angleški jezik
• Naslov: Quality control of automotive catalysts assembly with machine vision

Izvleček

This master thesis is looking into the quality control with machine vision, which is becoming an important segment in the assembly of catalytic converters. In this thesis, because of a specific converter, we upgraded the existing inspection algorithm in such a way that it will be able to divide the converter surface into two regions of interests. Each region should be inspected with different criteria. In the beginning of the thesis the catalytic converter and its functions are presented, as well as basics of quality control and image processing. Additionally, current inspection system and overview of the market solutions are described. The development of the new algorithm and the integration into the existing software are described in the second part of the thesis. Other avenues of research with MATLAB are also presented in this part. In the final part of the thesis, the experimenting with the new algorithm and the analysis of results is shown.

• Ključne besede: - machine vision - quality control - image segmentation - catalytic converter - obstructed channel algorithm