Segmentacija krvnih žila u slikama očne pozadine

master's thesis

• Title (english): Segmentation of Blood Vessels in the Images of the Ocular Background
• Author: Jakov Kruljac
• Mentor: Marko Subašić (mentor)
• Committee member: Marko Subašić (predsjednik povjerenstva)
• Committee member: Sven Lončarić (član povjerenstva)
• Committee member: Zvonko Kostanjčar (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-12, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U ovom radu obradila se problematika segmentacije krvnih žila iz očne pozadine. Objašnjeni se različite vrste neuronskih mreža korištene u predloženom modelu. Posebna pažnja posvetila se modelima koji mogu raditi s malim brojem podataka za učenje kao što je U-net. Također koristila se metoda generativnih suparničkih neuronskih mreža što je pozitivno utjecalo na rad U-neta. U radu je predloženi model uspješno testiran na problemu prepoznavanja malih krvnih žila koje nisu označene od strane doktora. Predloženi model prepoznaje gotovo sve žile označene od strane doktora 84% te dodatno pronalazi manje krvne žile koje doktori nisu označili zbog čega preciznost modela iznosi 34%.

Abstract (english)

This paper deals with the problem of segmentation of blood vessels from the retina. The different types of neural networks used in the proposed model are explained. Special attention has been paid to models that can work with a small amount of learning data such as U-net. The method of generative adversarial networks was also used, which had a positive effect on the operation of U-net. In this paper, the proposed model is successfully tested on the problem of recognizing small blood vessels that are not marked by a doctor. The proposed model recognizes almost all vessels marked by doctors 84% and additionally finds smaller blood vessels that were not marked by doctors, which is why the accuracy of the model is 34%.

• Keywords: krvne žile očne pozadine; generativne suparničke mreže; U-net; segmentacija krvih žila
• Keywords (english): blood vessels from retina; generative adversarial networks; U-net; blood vessels segmentation
• Language: croatian
• URN:NBN: urn:nbn:hr:168:783166
• 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: 2025-02-25 16:12:40