Usporedba metoda za mjerenje razlika između skupova vizualnih podataka u kontekstu primjene u robotici

master's thesis

• Title (english): Comparison of methods for measuring differences between visual datasets in the context of robotics applications
• Author: Antun Grozdek
• Mentor: Stjepan Bogdan (mentor)
• Committee member: Stjepan Bogdan (član povjerenstva)
• Committee member: Goran Vasiljević (član povjerenstva)
• Committee member: Frano Petric (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-04, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj diplomski rad analizira tehnike za mjerenje razlika u skupovima vizualnih podataka. Istraživanje je pokazalo da je najbolja tehnika za daljnja istraživanja Earth Mover’s Distance (EMD). EMD je zahtijevao prikaz slike, pa su u ovom radu predstavljene tri metode: RGB metoda, histogramska metoda i SIFT metoda. Usporedna analiza pokazala je da je vremenski najbolja RGB metoda, najveću točnost ima histogram, a najbolju ideju za budući razvoj ima SIFT metoda. Za sada niti jedna od ovih metoda nije spremna za korištenje u aplikacijama robotike, ali metoda histograma i metoda SIFT imaju perspektivu za to.

Abstract (english)

This thesis analyzes techniques for measuring differences in visual datasets. Investigation indicated that the best technique for further research is Earth Mover’s Distance
(EMD). EMD demanded image representation, so three methods of it are represented
in this thesis: the RGB method, the histogram method, and the SIFT method. Comparative analysis has shown that time-wise, the RGB method is the best, accuracy-wise,
the histogram is the best, and the best idea for future development is the SIFT method.
As of now, neither of these methods is ready for usage in robotics applications, but the
histogram method and the SIFT method have prospects for it.

• Keywords: skupovi vizualnih podataka; aplikacije robotike; Earth Mover’s Distance; vektori obilježja; RGB; histogram; SIFT; prikaz slike
• Keywords (english): visual datasets; robotics applications; Earth Mover’s Distance; feature vectors; RGB; histogram; SIFT; image representation
• Language: english
• URN:NBN: urn:nbn:hr:168:929733
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-10-11 08:58:58