Kompresija slik s pomočjo linearne regresije in delitve slike na ploske sektorje.

PRIMOŽIČ, GAŠPER (author), Kristan, Matej (mentor)

V diplomskem delu predlagamo novo metodo izgubnega ali neizgubnega stiskanja slik. Slike so v njihovi osnovni obliki, v računalniku predstavljene kot dvodimenzionalna polja intenzitet treh različnih barvnih kanalov. Barvni model slike spremenimo v takšnega, ki minimizira kovarianco med intenzitetami vseh barvnih kanalov, hkrati pa varianco maksimizira na enem samem kanalu. S pomočjo sektorizacije, posamezne barvne kanale razrežemo na sektorje, katerega intenzitete so sebi podobne. Intenzitete vseh stolpcev in vrstic dobljenih sektorjev aproksimiramo s pomočjo regresijskih funkcij prve stopnje. Tako za vsak sektor slike, namesto vseh intenzitet, ohranimo le koeficiente, ki opisujejo priležne premice in na ta način dosežemo kompresijo. Ker so lokacije intenzitet vrstice ali stolpca vnaprej znane, lahko postopek iskanja koeficientov tovrstnih premic okrajšamo. Prav tako pa metoda predstavlja tudi možnost paralelizacije korakov kompresije in dekompresije, saj se glavnina računanja izvaja z matrikami, za kar pa so moderni procesorji in grafične kartice specializirani. Predstavimo tudi evalvacijo predstavljene metode. S pomočjo numeričnih orodij ocenjevanja kakovosti rekonstrukcij in statistične analize pokažemo, da se prednosti metode izkažejo pri močno stisnjenih slikah, katere so po večini ploske.

• Language: Slovenian
• Keywords: kompresija, računalnik, linearna regresija
• Work type: Bachelor thesis/paper
• Organization: FRI - Faculty of Computer and Information Science
• Year: 2017
• PID: 20.500.12556/RUL-99047
• Publication date in RUL: 22.12.2017

PRIMOŽIČ, GAŠPER, 2017, Kompresija slik s pomočjo linearne regresije in delitve slike na ploske sektorje. [online]. Bachelor’s thesis. [Accessed 10 June 2025]. Retrieved from: https://repozitorij.uni-lj.si/IzpisGradiva.php?lang=eng&id=99047

Secondary language

• Language: English
• Title: Image compression with linear regression and image segmentation to flat sectors

Abstract

We propose a new approach for lossless and lossy image compression. Images are stored in computer in its primary form, as two dimensional fields of intensities of three color channels. We change the color model of an image to one which minimizes the covariance between the intensities of all channels, and maximizes the variance on a single one. With the help of sectorization, we split each of the color channels to sectors, whose intensities are self similar. We approximate the intensities of all sectors, by applying first degree regression functions, to all rows and columns of a sector. Compression is achieved by storing only the coefficients of the regressed functions and discarding the original intensities. The procedure of finding the regression function coefficients can be speeded up since the positions of the intensities in a sector are known beforehand. This method is also expandable for paralellisation of crucial steps in the algorithm, because the majority of the time consuming calculations are matrix-based. Modern day processors and graphics cards are made for these kinds of applications, which can drastically drop the compression and decompression times. Our proposed compression method is thoroughly analyzed. Numerical approaches and statistical analysis are utilized to show that the advantage of this method is with highly compressed flat images.

• Keywords: compression, computer, linear regression