A novel computer desktop image compression technology based on clustering algorithm

This paper proposes the new perspective on image compression algorithm with the assistance of the machine learning algorithms. Inspired from the literature review, with the continuous development of multimedia and the Internet application, and more and more high to the requirement of the image compr...

Full description

Saved in:
Bibliographic Details
Published in:RISTI : Revista Ibérica de Sistemas e Tecnologias de Informação no. E6; p. 295
Main Author: Wang, Xing
Format: Journal Article
Language:English
Portuguese
Published: Lousada AISTI (Iberian Association for Information Systems and Technologies) 01.08.2016
Associação Ibérica de Sistemas e Tecnologias de Informacao
Subjects:
Algorithms
Data compression
Data mining
Equipment and supplies
Image processing
Internet software
Machine learning
ISSN:1646-9895
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper proposes the new perspective on image compression algorithm with the assistance of the machine learning algorithms. Inspired from the literature review, with the continuous development of multimedia and the Internet application, and more and more high to the requirement of the image compression technology, it was found that have some static image compression standard has been difficult to meet certain requirements. We therefore propose the new idea on the image compression based on the wavelet analysis and clustering model. To conduct the reasonable experiment, we solve as follows. For the color image quality measure problem, because the color perception is closely related to the human visual characteristics, on a different color and color changes in different directions, and the human eye perception of sensitivity is not the same, so that the color measurement is a complicated process. This makes the color image quality evaluation than gray image is much more difficult. In order to simple, objective for the purpose of, still color image restoration using PSNR as quality evaluation, but we will list three component of color image PSNR, respectively. The experiment proves the effectiveness of the method. Keywords: Image processing, pattern recognition, data compression, clustering algorithm, desktop image
Bibliography:SourceType-Scholarly Journals-1
ObjectType-General Information-1
content type line 14
ISSN:1646-9895