prikaz prve stranice dokumenta Uklanjanje objekata sa slika pomoću dubokog učenja
  Download
PDF 2.1 MB
master's thesis
Uklanjanje objekata sa slika pomoću dubokog učenja
Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2023. urn:nbn:hr:200:547603
Varšava, Marko
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Software Engineering
Chair of Visual Computing

Institutional repository: Faculty of Electrical Engineering, Computer Science and Information Technology Osijek

Cite this document

Varšava, M. (2023). Uklanjanje objekata sa slika pomoću dubokog učenja (Master's thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek. Retrieved from https://urn.nsk.hr/urn:nbn:hr:200:547603

Varšava, Marko. "Uklanjanje objekata sa slika pomoću dubokog učenja." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2023. https://urn.nsk.hr/urn:nbn:hr:200:547603

Varšava, Marko. "Uklanjanje objekata sa slika pomoću dubokog učenja." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2023. https://urn.nsk.hr/urn:nbn:hr:200:547603

Varšava, M. (2023). 'Uklanjanje objekata sa slika pomoću dubokog učenja', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 08 May 2024, https://urn.nsk.hr/urn:nbn:hr:200:547603

Varšava M. Uklanjanje objekata sa slika pomoću dubokog učenja [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2023 [cited 2024 May 08] Available at: https://urn.nsk.hr/urn:nbn:hr:200:547603

M. Varšava, "Uklanjanje objekata sa slika pomoću dubokog učenja", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2023. Available at: https://urn.nsk.hr/urn:nbn:hr:200:547603

Metadata
TitleUklanjanje objekata sa slika pomoću dubokog učenja
Title (english)Removing objects from images with deep learning
AuthorMarko Varšava
MentorČaslav Livada (mentor)
Committee memberČaslav Livada (predsjednik povjerenstva)
Committee memberKrešimir Nenadić (član povjerenstva)
Committee memberRobert Šojo (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
(Department of Software Engineering)
(Chair of Visual Computing)
Osijek
Defense date and country2023-10-14, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Data Processing
Abstract
Ovaj diplomski rad istražuje kako kombinacija tehnika dubokog učenja segmentacija objekata sa slike i inpainting može ubrzati uklanjanje objekata sa slika. Na početku su objašnjenja postojeća rješenja za uklanjanje objekata sa slika. Za rješenje ovog problema koristi se model za semantičku segmentaciju SAM za generiranje maske, a za popunjavanje maskiranog dijela slike koristi se model EdgeConnect. Modeli su detaljnije objašnjeni. Rad je fokusiran na uklanjanju osoba s fotografija ulica, za trening su generirane maske silueta ljudi i preuzet skup podataka slika iz kojeg su izdvojene slike ulica i gradova. Model je dotreniran na predtreniranom modelu treniranom nad skupom podataka ParisStreetView.
Abstract (english)
This thesis explores how the combination of deep learning techniques for object segmentation from images and inpainting can expedite the removal of objects from pictures. At the outset, existing solutions for object removal from images are explained. To address this issue, the SAM model for semantic segmentation is used to generate a mask, and the EdgeConnect model is utilized for filling in the masked portion of the image. The models are explained in detail. The research is focused on the removal of people from street photographs. For training, silhouette masks of people were generated, and a dataset containing images was acquired, from which street and city images were extracted. The model was fine-tuned on a pre-trained model trained on the ParisStreetView dataset.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:547603
Study programmeTitle: Graduate University Study Programme in Computer Engineering
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2023-10-30 13:15:07