On spatiochromatic features in natural images statistics

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this communication, we show that two simple assumptions on the covariances matrices of color images, namely stationarity and commutativity, can explain the observed shape of decorrelated spatiochromatic elements (bases obtained by PCA) of natural color images. The validity of these assumptions is tested on a large database of RAW images. Our experiments also show that the spatiochromatic covariance decays exponentially with the spatial distance between pairs of pixels and not as a power law as it is commonly assumed.

Original languageEnglish
Title of host publicationImage Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings
EditorsVittorio Murino, Vittorio Murino, Enrico Puppo
PublisherSpringer Verlag
Pages46-55
Number of pages10
ISBN (Print)9783319232331
DOIs
Publication statusPublished - 1 Jan 2015
Externally publishedYes
Event18th International Conference on Image Analysis and Processing, ICIAP 2015 - Genoa, Italy
Duration: 7 Sept 201511 Sept 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9280
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference18th International Conference on Image Analysis and Processing, ICIAP 2015
Country/TerritoryItaly
CityGenoa
Period7/09/1511/09/15

Fingerprint

Dive into the research topics of 'On spatiochromatic features in natural images statistics'. Together they form a unique fingerprint.

Cite this