On Decomposing Object Appearance using PCA and Wavelet bases with Applications to Image Segmentation

Mikkel Bille Stegmann; Søren Forchhammer; Hans Joachim Werner

On Decomposing Object Appearance using PCA and Wavelet bases with Applications to Image Segmentation

Mikkel Bille Stegmann, Søren Forchhammer, Hans Joachim Werner (Editor)

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

56 Downloads (Pure)

Abstract

Generative models capable of synthesising complete object images have over the past few years proven their worth when interpreting images. Due to the recent development of computational machinery it has become feasible to model the variation of image intensities and landmark positions over the complete object surface using principal component analysis. This typically involves matrices with a few thousands and up to 100.000+ rows. This paper demonstrates applications of such models applied on colour images of human faces and cardiac magnetic resonance images. Further, we devise methods for alleviating the obvious computational and storage requirements of these large models by means of truncated wavelet bases.

Original language	English
Title of host publication	MATRIX'02, Eleventh International Workshop on Matrices and Statistics
Number of pages	21
Publisher	Informatics and Mathematical Modelling, Technical University of Denmark, DTU
Publication date	2002
Publication status	Published - 2002
Event	Eleventh International Workshop on Matrices and Statistics - Richard Petersens Plads, Building 321, DK-2800 Kgs. Lyngby, Kgs. Lyngby, Denmark Duration: 29 Aug 2002 → 31 Aug 2002 Conference number: 11

Conference

Conference	Eleventh International Workshop on Matrices and Statistics
Number	11
Location	Richard Petersens Plads, Building 321, DK-2800 Kgs. Lyngby
Country/Territory	Denmark
City	Kgs. Lyngby
Period	29/08/2002 → 31/08/2002

Access to Document

Imm1017

http://www.imm.dtu.dk/pubdb/p.php?1017

OpenUrl availability

Full text

Cite this

@inproceedings{312b323580b54b6e88735988fa2659f3,

title = "On Decomposing Object Appearance using PCA and Wavelet bases with Applications to Image Segmentation",

abstract = "Generative models capable of synthesising complete object images have over the past few years proven their worth when interpreting images. Due to the recent development of computational machinery it has become feasible to model the variation of image intensities and landmark positions over the complete object surface using principal component analysis. This typically involves matrices with a few thousands and up to 100.000+ rows. This paper demonstrates applications of such models applied on colour images of human faces and cardiac magnetic resonance images. Further, we devise methods for alleviating the obvious computational and storage requirements of these large models by means of truncated wavelet bases.",

author = "Stegmann, {Mikkel Bille} and S{\o}ren Forchhammer and Werner, {Hans Joachim}",

year = "2002",

language = "English",

booktitle = "MATRIX'02, Eleventh International Workshop on Matrices and Statistics",

publisher = "Informatics and Mathematical Modelling, Technical University of Denmark, DTU",

note = "Eleventh International Workshop on Matrices and Statistics, MATRIX'02 ; Conference date: 29-08-2002 Through 31-08-2002",

}

Stegmann, MB, Forchhammer, S & Werner, HJ (ed.) 2002, On Decomposing Object Appearance using PCA and Wavelet bases with Applications to Image Segmentation. in MATRIX'02, Eleventh International Workshop on Matrices and Statistics. Informatics and Mathematical Modelling, Technical University of Denmark, DTU, Eleventh International Workshop on Matrices and Statistics, Kgs. Lyngby, Denmark, 29/08/2002. <http://www.imm.dtu.dk/pubdb/p.php?1017>

On Decomposing Object Appearance using PCA and Wavelet bases with Applications to Image Segmentation. / Stegmann, Mikkel Bille; Forchhammer, Søren; Werner, Hans Joachim (Editor).
MATRIX'02, Eleventh International Workshop on Matrices and Statistics. Informatics and Mathematical Modelling, Technical University of Denmark, DTU, 2002.

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - On Decomposing Object Appearance using PCA and Wavelet bases with Applications to Image Segmentation

AU - Stegmann, Mikkel Bille

AU - Forchhammer, Søren

A2 - Werner, Hans Joachim

N1 - Conference code: 11

PY - 2002

Y1 - 2002

N2 - Generative models capable of synthesising complete object images have over the past few years proven their worth when interpreting images. Due to the recent development of computational machinery it has become feasible to model the variation of image intensities and landmark positions over the complete object surface using principal component analysis. This typically involves matrices with a few thousands and up to 100.000+ rows. This paper demonstrates applications of such models applied on colour images of human faces and cardiac magnetic resonance images. Further, we devise methods for alleviating the obvious computational and storage requirements of these large models by means of truncated wavelet bases.

AB - Generative models capable of synthesising complete object images have over the past few years proven their worth when interpreting images. Due to the recent development of computational machinery it has become feasible to model the variation of image intensities and landmark positions over the complete object surface using principal component analysis. This typically involves matrices with a few thousands and up to 100.000+ rows. This paper demonstrates applications of such models applied on colour images of human faces and cardiac magnetic resonance images. Further, we devise methods for alleviating the obvious computational and storage requirements of these large models by means of truncated wavelet bases.

M3 - Article in proceedings

BT - MATRIX'02, Eleventh International Workshop on Matrices and Statistics

PB - Informatics and Mathematical Modelling, Technical University of Denmark, DTU

T2 - Eleventh International Workshop on Matrices and Statistics

Y2 - 29 August 2002 through 31 August 2002

ER -

On Decomposing Object Appearance using PCA and Wavelet bases with Applications to Image Segmentation

Abstract

Conference

Access to Document

OpenUrl availability

Fingerprint

Cite this