Stiskanje slik : delo diplomskega seminarja

Golob, Klara (author), Žagar, Emil (mentor)

Stiskanje slik je danes nepogrešljiv del vsakodnevnega življenja ljudi. Ločimo dva postopka, stiskanje z izgubami in stiskanje brez izgub. Eden izmed najbolj uporabnih algoritmov za stiskanje je JPEG algoritem, ki temelji na stiskanju z izgubami. Za učinkovito stiskanje uporabi diskretno kosinusno transformacijo, ki pretvori sliko v frekvenčno domeno, kjer lažje odstrani nepomembne dele. Kljub temu, da se ponavadi nekaj podatkov pri poteku JPEG algoritma izgubi, lahko z inverznim algoritmom rekonstruiramo približek originalne slike. Rekonstruirana slika ni čisto enaka prvotni, ji je pa toliko podobna, da ponavadi s človeškim očesom ne zaznamo razlike.

• Language: Slovenian
• Keywords: stiskanje slik, JPEG standard, osnovni JPEG algoritem, diskretna kosinusna transformacija, kvantizacija, cik-cak skeniranje, Huffmanovo kodiranje, razširjanje slik
• Work type: Final seminar paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FMF - Faculty of Mathematics and Physics
• Year: 2020
• PID: 20.500.12556/RUL-120196
• UDC: 519.6
• COBISS.SI-ID: 58404867
• Publication date in RUL: 17.09.2020

GOLOB, Klara, 2020, Stiskanje slik : delo diplomskega seminarja [online]. Bachelor’s thesis. [Accessed 15 August 2025]. Retrieved from: https://repozitorij.uni-lj.si/IzpisGradiva.php?lang=eng&id=120196

Secondary language

• Language: English
• Title: Image compression

Abstract

Image compression is nowadays an indispensable part of our daily lives. It falls into two groups, lossy compression and lossless compression. One of the most useful image compression algorithms is JPEG algorithm, which belongs to the lossy compression group. For efficient compression, JPEG algorithm uses the discrete cosine transformation, which transforms the image into a frequency domain, that we can more easily remove insignificant parts of the image. Despite the fact that the data is lost during the JPEG algorithm, we can retrieve this data in the reverse order of compression and reconstruct the approximation of original image. Reconstructed image is not exactly the same as the original, but it is so similar that we usually do not notice the difference with our human eye.

• Keywords: image compression, JPEG standard, main JPEG algorithm, discrete cosine transform, quantization, cik-cak scanning, Huffman coding, image decompression