High-pass filters give histograms with positive kurtosis

Eric Clarkson; Harrison H. Barrett

doi:10.1364/OL.26.001253

High-pass filters give histograms with positive kurtosis

Eric Clarkson
, Harrison H. Barrett

Research output: Contribution to journal › Article › peer-review

2 Scopus citations

Abstract

If we filter an ensemble of natural images with a high-pass filter, the output is a random variable. The probability density for this variable will, under fairly general assumptions, be a Gaussian scale mixture. The kurtosis for this type of distribution is always positive.

Original language	English (US)
Pages (from-to)	1253-1255
Number of pages	3
Journal	Optics letters
Volume	26
Issue number	16
DOIs	https://doi.org/10.1364/OL.26.001253
State	Published - Aug 15 2001
Externally published	Yes

ASJC Scopus subject areas

Atomic and Molecular Physics, and Optics

Access to Document

10.1364/OL.26.001253

Cite this

@article{06803c69a096493d9293a0e1edd73170,

title = "High-pass filters give histograms with positive kurtosis",

abstract = "If we filter an ensemble of natural images with a high-pass filter, the output is a random variable. The probability density for this variable will, under fairly general assumptions, be a Gaussian scale mixture. The kurtosis for this type of distribution is always positive.",

author = "Eric Clarkson and Barrett, \{Harrison H.\}",

year = "2001",

month = aug,

day = "15",

doi = "10.1364/OL.26.001253",

language = "English (US)",

volume = "26",

pages = "1253--1255",

journal = "Optics letters",

issn = "0146-9592",

publisher = "Optica Publishing Group (formerly OSA)",

number = "16",

}

TY - JOUR

T1 - High-pass filters give histograms with positive kurtosis

AU - Clarkson, Eric

AU - Barrett, Harrison H.

PY - 2001/8/15

Y1 - 2001/8/15

N2 - If we filter an ensemble of natural images with a high-pass filter, the output is a random variable. The probability density for this variable will, under fairly general assumptions, be a Gaussian scale mixture. The kurtosis for this type of distribution is always positive.

AB - If we filter an ensemble of natural images with a high-pass filter, the output is a random variable. The probability density for this variable will, under fairly general assumptions, be a Gaussian scale mixture. The kurtosis for this type of distribution is always positive.

UR - https://www.scopus.com/pages/publications/0141615917

UR - https://www.scopus.com/pages/publications/0141615917#tab=citedBy

U2 - 10.1364/OL.26.001253

DO - 10.1364/OL.26.001253

M3 - Article

C2 - 18049577

AN - SCOPUS:0141615917

SN - 0146-9592

VL - 26

SP - 1253

EP - 1255

JO - Optics letters

JF - Optics letters

IS - 16

ER -

High-pass filters give histograms with positive kurtosis

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this