Metode rastriranja crno bijele slike

master's thesis

• Author: Nika Sviben
• Mentor: Ante Poljičak (mentor)
• Committee member: Lidija Mandić (član povjerenstva)
• Committee member: Ante Poljičak (član povjerenstva)
• Committee member: Maja Strgar Kurečić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Graphic Arts; Zagreb
• Defense date and country: 2016-9-27, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Graphic Technology

Abstract

Rastriranje nije jedan, strogo definiran fotomehanički proces nego opisuje dva procesa. U prvom se kontinuirani tonovi pretvaraju u seriju točaka. U drugom se pomoću raznih metoda mehaničkog otiskivanja otiskuju slike koje simuliraju kontinuirani ton. Osnovne metode digitalnog rastriranja slike su: jednotonska segmentacija, tablično rastriranje, rastriranje matricom nivoa i rastriranje difuzijom greške. HVS metode ispitivanja kvalitete slike su objektivne, kvalitativne procjene slike bazirane na ljudskom vizualnom sustavu. Postoji više metoda baziranih na HVS-u. WSNR se primjenjuje na slike na način da se prvo izračuna razlika slike kontinuiranog tona i obrađene slike te se pogreška ponderira s odazivom frekvencije HVS-a danim sa CSF-om.

U radu je ispitano 1000 crno-bijelih slika rastriranih koristeći 3 metode: tablično rastriranje, rastriranje matricom nivoa i rastriranje difuzijom greške. Rastriranje je provedeno putem programa MATLAB. Nakon rastriranja ispitana je kvaliteta dobivenih slika putem WSNR-a te su uspoređeni rezultati. Rezultati su pokazali kako metoda difuzije greške daje najbolje rezultate rastriranja, matrično rastriranje daje nešto lošije rezultate, a tablično rastriranje, neovisno o rasporedu rasterskih elemenata daje najlošije rezultate. Također, pokazalo se kako je moguće poboljšati kvalitetu slike rastrirane metodom tabličnog rastriranja promjenom rasporeda rasterskih elemenata.

Abstract (english)

Halftoning is not one strictly defined photomechanical process. It describes two processes. First, continuous tones are converted into a series of dots. Than, images are printed by using various methods of mechanical printing. Printed images simulate continuous tone. Halftoning methods are generally divided according to the structure of the elements into two groups: amplitude modulated (AM) and frequency modulated (FM) halftoning. Basic methods of digital halftoning are monotone segmentation, table halftoning, threshold matrix and error diffusion. HVS test methods are objective image quality measurements based on the human visual system. There are several methods based on the HVS. WSNR is applied to the image by calculating the difference between the continuous tone and the processed image and than weighting the error of the response frequency of HVS with the given CSF.

1,000 black-and-white images where halftoned using three methods: table halftoning, threshold matrix and error diffusion. Halftoning was performed by the program MATLAB. After halftoning, quality measurement of halftoned images was conducted using WSNR and results were compared. Results show that error diffusion method produces halftoned images of best quality, threshold matrix produces halftoned images of lesser quality and table halftoning produces images of worst quality. It is possible to produces better quality images by changing the layout of the elements in the table.

• Keywords: rastriranje; tablično rastriranje; rastriranje matricom nivoa; difuzija greške; HVS
• Keywords (english): halftoning; table halftoning; threshold matrix; error diffusion; HVS
• Language: croatian
• URN:NBN: urn:nbn:hr:216:193934
• Study programme: Title: Graphic technology; specializations in: Graphic technology, Design of graphic products; Course: Graphic technology; Study programme type: university; Study level: graduate; Academic / professional title: sveučilišni/a magistar/magistra inženjer/inženjerka grafičke tehnologije (sveučilišni/a magistar/magistra inženjer/inženjerka grafičke tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2024-04-18 21:27:54