Super rezolucija slika tehnikama dubokih neuronskih mreža

master's thesis

• Author: Tin Mavračić
• Mentor: Tomislav Šmuc (mentor)
• Committee member: Tomislav Šmuc (predsjednik povjerenstva)
• Committee member: Robert Manger (član povjerenstva)
• Committee member: Saša Singer (član povjerenstva)
• Committee member: Goran Igaly (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Mathematics); Zagreb
• Defense date and country: 2018-11-27, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Mathematics

Abstract

Računalni vid jedno je od standardnih područja strojnog učenja i umjetne inteligencije. Tipični zadaci računalnog vida mogu se svrstati u dohvat, obradu, analizu i razumijevanje slika. Super rezolucija slika jedan je od postupaka predobrade, kojim se nastoji povećati postojeća rezolucija slike. Često se koristi u obradi satelitskih i medicinskih slika, te kao postupak u različitim multimedijskim aplikacijama, kako bi se riješili problemi zbog prisutnosti šuma, zamagljenih i degradiranih područja. U okviru ovog rada istraženi su pristupi za rješavanje problema super rezolucije uz širok pregled područja, te je implementirano rješenje super rezolucije pomoću dubokih neuronskih mreža.

Abstract (english)

Computer vision is one of the standard areas of machine learning and artificial intelligence. Typical computer vision tasks include extraction, processing, analysing and understanding images. Image super resolution is one of the preprocessing steps by which one wishes to enlarge existing image resolution. It is often used in satellite and medical images, but also as processing step in multimedia applications to resolve problems caused by noise, blur, glare and other image degrading processes. Wide area of super resolution methods have been examined within this master’s thesis, and an approach based on deep neural networks was implemented.

• Keywords: računalni vid; strojno učenje; umjetna inteligencija; super rezolucija slika; duboke neuronske mreže
• Keywords (english): computer vision; machine learning; artificial intelligence; image super resolution; deep neural networks
• Language: croatian
• URN:NBN: urn:nbn:hr:217:618444
• Study programme: Title: Computer Science and Mathematics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra računarstva i matematike (magistar/magistra računarstva i matematike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2019-06-07 11:02:18