Directly invertible nonlinear divisive normalization pyramid for image representation

Roberto Valerio; Eero P. Simoncelli; Rafael Navarro

doi:10.1007/978-3-540-39798-4_42

Directly invertible nonlinear divisive normalization pyramid for image representation

Roberto Valerio, Eero P. Simoncelli, Rafael Navarro

Neural Science

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Abstract

We present a multiscale nonlinear image representation that permits an efficient coding of natural images. The input image is first decomposed into a set of subbands at multiple scales and orientations using near-orthogonal symmetric quadrature mirror filters. This is followed by a nonlinear "divisive normalization" stage, in which each linear coefficient is divided by a value computed from a small set of neighboring coefficients in space, orientation and scale. This neighborhood is chosen to allow this nonlinear operation to be efficiently inverted. The parameters of the normalization operation are optimized in order to maximize the independence of the normalized responses for natural images. We demonstrate the near-independence of these nonlinear responses, and suggest a number of applications for which this representation should be well suited.

Original language	English (US)
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Editors	Narciso Garcia, Luis Salgado, Jose M. Martinez
Publisher	Springer Verlag
Pages	331-340
Number of pages	10
ISBN (Print)	3540200819, 9783540200819
DOIs	https://doi.org/10.1007/978-3-540-39798-4_42
State	Published - 2003

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	2849
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-540-39798-4_42

Cite this

Valerio, R., Simoncelli, E. P., & Navarro, R. (2003). Directly invertible nonlinear divisive normalization pyramid for image representation. In N. Garcia, L. Salgado, & J. M. Martinez (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 331-340). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2849). Springer Verlag. https://doi.org/10.1007/978-3-540-39798-4_42

Directly invertible nonlinear divisive normalization pyramid for image representation. / Valerio, Roberto; Simoncelli, Eero P.; Navarro, Rafael.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). ed. / Narciso Garcia; Luis Salgado; Jose M. Martinez. Springer Verlag, 2003. p. 331-340 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2849).

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Valerio, R, Simoncelli, EP & Navarro, R 2003, Directly invertible nonlinear divisive normalization pyramid for image representation. in N Garcia, L Salgado & JM Martinez (eds), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2849, Springer Verlag, pp. 331-340. https://doi.org/10.1007/978-3-540-39798-4_42

Valerio R, Simoncelli EP, Navarro R. Directly invertible nonlinear divisive normalization pyramid for image representation. In Garcia N, Salgado L, Martinez JM, editors, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Springer Verlag. 2003. p. 331-340. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-540-39798-4_42

Valerio, Roberto ; Simoncelli, Eero P. ; Navarro, Rafael. / Directly invertible nonlinear divisive normalization pyramid for image representation. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). editor / Narciso Garcia ; Luis Salgado ; Jose M. Martinez. Springer Verlag, 2003. pp. 331-340 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{5081e9e611604a6c9620aa972eb0e7a2,

title = "Directly invertible nonlinear divisive normalization pyramid for image representation",

abstract = "We present a multiscale nonlinear image representation that permits an efficient coding of natural images. The input image is first decomposed into a set of subbands at multiple scales and orientations using near-orthogonal symmetric quadrature mirror filters. This is followed by a nonlinear {"}divisive normalization{"} stage, in which each linear coefficient is divided by a value computed from a small set of neighboring coefficients in space, orientation and scale. This neighborhood is chosen to allow this nonlinear operation to be efficiently inverted. The parameters of the normalization operation are optimized in order to maximize the independence of the normalized responses for natural images. We demonstrate the near-independence of these nonlinear responses, and suggest a number of applications for which this representation should be well suited.",

author = "Roberto Valerio and Simoncelli, {Eero P.} and Rafael Navarro",

year = "2003",

doi = "10.1007/978-3-540-39798-4_42",

language = "English (US)",

isbn = "3540200819",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "331--340",

editor = "Narciso Garcia and Luis Salgado and Martinez, {Jose M.}",

booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - CHAP

T1 - Directly invertible nonlinear divisive normalization pyramid for image representation

AU - Valerio, Roberto

AU - Simoncelli, Eero P.

AU - Navarro, Rafael

PY - 2003

Y1 - 2003

N2 - We present a multiscale nonlinear image representation that permits an efficient coding of natural images. The input image is first decomposed into a set of subbands at multiple scales and orientations using near-orthogonal symmetric quadrature mirror filters. This is followed by a nonlinear "divisive normalization" stage, in which each linear coefficient is divided by a value computed from a small set of neighboring coefficients in space, orientation and scale. This neighborhood is chosen to allow this nonlinear operation to be efficiently inverted. The parameters of the normalization operation are optimized in order to maximize the independence of the normalized responses for natural images. We demonstrate the near-independence of these nonlinear responses, and suggest a number of applications for which this representation should be well suited.

AB - We present a multiscale nonlinear image representation that permits an efficient coding of natural images. The input image is first decomposed into a set of subbands at multiple scales and orientations using near-orthogonal symmetric quadrature mirror filters. This is followed by a nonlinear "divisive normalization" stage, in which each linear coefficient is divided by a value computed from a small set of neighboring coefficients in space, orientation and scale. This neighborhood is chosen to allow this nonlinear operation to be efficiently inverted. The parameters of the normalization operation are optimized in order to maximize the independence of the normalized responses for natural images. We demonstrate the near-independence of these nonlinear responses, and suggest a number of applications for which this representation should be well suited.

UR - http://www.scopus.com/inward/record.url?scp=24644510240&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=24644510240&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-39798-4_42

DO - 10.1007/978-3-540-39798-4_42

M3 - Chapter

AN - SCOPUS:24644510240

SN - 3540200819

SN - 9783540200819

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 331

EP - 340

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

A2 - Garcia, Narciso

A2 - Salgado, Luis

A2 - Martinez, Jose M.

PB - Springer Verlag

ER -

Directly invertible nonlinear divisive normalization pyramid for image representation

Abstract

Publication series

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this