Izboljšana metoda stiskanja risanih slik

magistrsko delo

Aljaž Jeromel (Author), Borut Žalik (Mentor)

Abstract

V magistrskem delu opišemo izboljšave algoritma stiskanja risanih slik, ki smo ga predstavili v diplomski nalogi. Z uporabo osemsmerne povezanosti pikslov, združevanjem verižnih kod in optimizacijo zapisa RLE izboljšamo učinkovitost stiskanja, medtem ko z uporabo gručenja in uteženega združevanja območij uravnavamo kakovost rekonstrukcije slike. Pokažemo, da izboljšan algoritem risane slike stisne učinkoviteje, kot algoritem brez izboljšav, prav tako pa bolje ohrani podrobnosti v slikah. Učinkovitost izboljšanega algoritma primerjamo tudi s sorodnimi metodami stiskanja risanih slik, algoritmom PNG in standardom JPEG. Ugotovili smo, da učinkovitost predstavljenega algoritma na risanih slikah večkratno presega učinkovitosti vseh primerjanih metod.

Keywords

stiskanje slik;verižne kode;gručenje;transformacije nizov;magistrske naloge;

Data

Language:	Slovenian
Year of publishing:	2019
Typology:	2.09 - Master's Thesis
Organization:	UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher:	A. Jeromel
UDC:	528.852:004.932.1(043.2)
COBISS:	22340886
Views:	846
Downloads:	59
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Improved cartoon image compression method
Secondary abstract:	In this thesis, improvements of cartoon image compression algorithm, presented in our BsC thesis work, are described. The compression efficiency is improved by using the 8-connectivity, chain code concatenation, and RLE optimization, while clustering and weighted region merging improve the quality of the reconstructed images. It is shown by experiments, that the improved algorithm achieves better compression efficiency than the original one, despite better retaining small details in cartoon images. The compression efficiency was compared with the related methods, the algorithm PNG, and the JPEG standard. The compression efficiency, achieved by the presented algorithm, outperforms the compared methods by a few times.
Secondary keywords:	image compression;chain codes;clustering;string transformations;
URN:	URN:SI:UM:
Type (COBISS):	Master's thesis/paper
Embargo end date (OpenAIRE):	2021-04-19
Thesis comment:	Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Računalništvo in informacijske tehnologije
Pages:	VIII, 35 f.
ID:	11115488