Evaluacija objektivnih metoda za ocjenu kvalitete slike

undergraduate thesis

• Title (english): Evaluation of Objective Methods for Picture Quality Assessment
• Author: Domagoj Penava
• Mentor: Mislav Grgić (mentor)
• Committee member: Mislav Grgić (predsjednik povjerenstva)
• Committee member: Sonja Grgić (član povjerenstva)
• Committee member: Tomislav Kos (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

Objektivne metode za procjenu kvalitete slike općenito se dijele na dvije vrste: metode temeljenje na matematičkim modelima u koje spadaju PSNR, SSIM i UQI, te metode temeljenje na karakteristikama ljudskog vizualnog sustava, kao što su VSNR i WSNR.
Svaka od ovih pet metoda za vrednovanje kvalitete slike isprobane su na trima bazama slika: tid2008, csiq i live te su uspoređeni koeficijenti korelacije vrijednosti mjere te subjektivnih ocjena iz baze. Za izračunavanje korelacije iskorišteni su Spearmanov i Kendallov koeficijent korelacije.
Na dvjema od tri baze: tid2008 i live mjera VSNR pokazala je najbolju korelaciju sa subjektivnim ocjenama iz baze, dok su koeficijenti korelacije za sve mjere kod baze csiq bili podjednako jednaki. Na temelju rezultata iz baza pokazuje se da je mjera VSNR najvjerodostojnija, te se nju preporučuje koristiti.

Abstract (english)

Objective methods for image quality assessment are generally divided into two types: methods based on mathematical models, which include PSNR, SSIM and UQI, and methods based on the characteristics of the human visual system, such as VSNR and WSNR.
Each of these five methods for evaluating image quality were tested on three image databases: tid2008, csiq and live, and subsequently the correlation coefficients of the measured values and subjective ratings from the database were compared. Spearman's and Kendall's correlation coefficients were used to calculate the correlation.
VSNR showed the best correlation with subjective ratings when tested on two of the three image databases: tid2008 and live, while the correlation coefficients for all measures when tested on the csiq image database were all fairly equal. Based on the results from the databases, VSNR has shown itself to be the most reliable measure, and its use is recommended.

• Keywords: objektivne metode; procjena kvalitete slike; ljudski vizualni sustav; PSNR; SSIM; UQI; VSNR; WSNR; koeficijent korelacije; baza slika
• Keywords (english): Objective Methods; Image Quality Assessment; Human Visual System; PSNR; SSIM; UQI; VSNR; WSNR; Correlation Coefficient; Image Database
• Language: croatian
• URN:NBN: urn:nbn:hr:168:253760
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-07-13 10:26:33