prikaz prve stranice dokumenta Ocjena kvalitete video signala zasnovana na strojnom učenju
  Download
PDF 4.48 MB
master's thesis
Ocjena kvalitete video signala zasnovana na strojnom učenju
Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2023. urn:nbn:hr:200:170956
Ćosić, Tomislav
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Communications
Chair of Multimedia Systems and Digital Television

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

Cite this document

Ćosić, T. (2023). Ocjena kvalitete video signala zasnovana na strojnom učenju (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:200:170956

Ćosić, Tomislav. "Ocjena kvalitete video signala zasnovana na strojnom učenju." 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:200:170956

Ćosić, Tomislav. "Ocjena kvalitete video signala zasnovana na strojnom učenju." 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:200:170956

Ćosić, T. (2023). 'Ocjena kvalitete video signala zasnovana na strojnom učenju', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 05 November 2024, https://urn.nsk.hr/urn:nbn:hr:200:170956

Ćosić T. Ocjena kvalitete video signala zasnovana na strojnom učenju [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2023 [cited 2024 November 05] Available at: https://urn.nsk.hr/urn:nbn:hr:200:170956

T. Ćosić, "Ocjena kvalitete video signala zasnovana na strojnom učenju", 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:200:170956

Metadata
TitleOcjena kvalitete video signala zasnovana na strojnom učenju
Title (english)Video quality assessment based on machine learning
AuthorTomislav Ćosić
MentorMario Vranješ (mentor)
Committee memberMario Vranješ (predsjednik povjerenstva)
Committee memberMarijan Herceg (član povjerenstva)
Committee memberRatko Grbić (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
(Department of Communications)
(Chair of Multimedia Systems and Digital Television)
Osijek
Defense date and country2023-09-25, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Electrical Engineering
Telecommunications and Informatics
Abstract
U sklopu ovog diplomskog rada izrađen je algoritam za ocjenjivanje video signala zasnovan na strojnom učenju. Opisano je što su to subjektivna i objektivna procjena kvalitete video signala i na koji način se provode. Osim toga, dan je pregled postojećih algoritama za ocjenjivanje video signala zasnovanih na egzaktnim računalnim algoritmima te nekih postojećih algoritama za ocjenjivanje video signala zasnovanih na strojnom učenju. Nadalje, opisane su korištene baze video signala koje sadrže H.264, H.265, MPEG2, MPEG4 izobličenja i izobličenja nastala smrzavanjem, određenim % gubitka paketa pri prijenosu IP i bežičnim mrežama. Predložena su dva modela zasnovana na strojnom učenju koja rješavaju problem ocjenjivanja video signala bez potrebe za referentnim signalom. Za realizaciju rješenja korištena je Tensorflow biblioteka za Python programski jezik. Prikazani su rezultati testiranja za predložene modele koji su uspoređeni s rezultatima PSNR, SSIM, MS-SSIM i NIQE metrika. Predloženi modeli točnije predviđaju ocjene na testnom skupu nego objektivne metrike s kojima su uspoređeni za prvi način podjele podataka. Za drugi način podjele podataka Model 1 postiže bolje rezultate u odnosu na Model 2, no u odnosu na objektivne metrike s kojima je uspoređivan postiže bolje rezultate samo za H.264 tip izobličenja i izobličenja uzrokovana kompresijom.
Abstract (english)
Within this master's thesis, an algorithm for assessing video signals based on machine learning was developed. It describes the concepts of subjective and objective quality assessment of video signals and the methods used for conducting them. Additionally, an overview of existing algorithms for video signal assessment based on exact computational algorithms, as well as some existing algorithms based on machine learning, is provided. Furthermore, the used video signal databases containing H.264, H.265, MPEG2, MPEG4 distortions, and freeze distortions, as well as % of packet loss in IP and wireless networks, are described. Two machine learning-based models are proposed to address the problem of video signal assessment without the need for a reference signal. The TensorFlow library for the Python programming language was used for implementing the solutions. The testing results for the proposed models are presented and compared with the results of the PSNR, SSIM, MS-SSIM, and NIQE metrics. The proposed models achieve more accurate predictions on the test set compared to the objective metrics they were compared to for the first data partition method. For the second data partition method, Model 1 performs better than Model 2, but it achieves better results compared to the objective metrics only for H.264 distortion type and compression-induced distortions.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:170956
Study programmeTitle: Graduate University Study Programme in Electrical Engineering; specializations in: Communications and Informatics, Power Engineering
Course: Communications and Informatics
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike (magistar/magistra inženjer/inženjerka elektrotehnike)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2023-10-10 11:22:25