Higher order singular value decomposition of tensors for fusion of registered images

Michael G. Thomason, Jens Gregor

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

This paper describes a computational method using tensor math for higher order singular value decomposition (HOSVD) of registered images. Tensor decomposition is a rigorous way to expose structure embedded in multidimensional datasets. Given a dataset of registered 2-D images, the dataset is represented in tensor format and HOSVD of the tensor is computed to obtain a set of 2-D basis images. The basis images constitute a linear decomposition of the original dataset. HOSVD is data-driven and does not require the user to select parameters or assign thresholds. A specific application uses the basis images for pixel-level fusion of registered images into a single image for visualization. The fusion is optimized with respect to a measure of mean squared error. HOSVD and image fusion are illustrated empirically with four real datasets: (1) visible and infrared data of a natural scene, (2) MRI and x ray CT brain images, and in nondestructive testing (3) x ray, ultrasound, and eddy current images, and (4) x ray, ultrasound, and shearography images.

Original languageEnglish (US)
Article number013023
JournalJournal of Electronic Imaging
Volume20
Issue number1
DOIs
StatePublished - Jan 1 2011

Fingerprint

Singular value decomposition
Tensors
fusion
tensors
decomposition
X rays
Ultrasonics
Image fusion
Computational methods
Eddy currents
Nondestructive examination
Magnetic resonance imaging
Brain
Visualization
Pixels
Infrared radiation
shearography
x rays
eddy currents
format

All Science Journal Classification (ASJC) codes

  • Atomic and Molecular Physics, and Optics
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Higher order singular value decomposition of tensors for fusion of registered images. / Thomason, Michael G.; Gregor, Jens.

In: Journal of Electronic Imaging, Vol. 20, No. 1, 013023, 01.01.2011.

Research output: Contribution to journalArticle

@article{e79c9d67713943eb9b94654d87a3c825,
title = "Higher order singular value decomposition of tensors for fusion of registered images",
abstract = "This paper describes a computational method using tensor math for higher order singular value decomposition (HOSVD) of registered images. Tensor decomposition is a rigorous way to expose structure embedded in multidimensional datasets. Given a dataset of registered 2-D images, the dataset is represented in tensor format and HOSVD of the tensor is computed to obtain a set of 2-D basis images. The basis images constitute a linear decomposition of the original dataset. HOSVD is data-driven and does not require the user to select parameters or assign thresholds. A specific application uses the basis images for pixel-level fusion of registered images into a single image for visualization. The fusion is optimized with respect to a measure of mean squared error. HOSVD and image fusion are illustrated empirically with four real datasets: (1) visible and infrared data of a natural scene, (2) MRI and x ray CT brain images, and in nondestructive testing (3) x ray, ultrasound, and eddy current images, and (4) x ray, ultrasound, and shearography images.",
author = "Thomason, {Michael G.} and Jens Gregor",
year = "2011",
month = "1",
day = "1",
doi = "10.1117/1.3563592",
language = "English (US)",
volume = "20",
journal = "Journal of Electronic Imaging",
issn = "1017-9909",
publisher = "SPIE",
number = "1",

}

TY - JOUR

T1 - Higher order singular value decomposition of tensors for fusion of registered images

AU - Thomason, Michael G.

AU - Gregor, Jens

PY - 2011/1/1

Y1 - 2011/1/1

N2 - This paper describes a computational method using tensor math for higher order singular value decomposition (HOSVD) of registered images. Tensor decomposition is a rigorous way to expose structure embedded in multidimensional datasets. Given a dataset of registered 2-D images, the dataset is represented in tensor format and HOSVD of the tensor is computed to obtain a set of 2-D basis images. The basis images constitute a linear decomposition of the original dataset. HOSVD is data-driven and does not require the user to select parameters or assign thresholds. A specific application uses the basis images for pixel-level fusion of registered images into a single image for visualization. The fusion is optimized with respect to a measure of mean squared error. HOSVD and image fusion are illustrated empirically with four real datasets: (1) visible and infrared data of a natural scene, (2) MRI and x ray CT brain images, and in nondestructive testing (3) x ray, ultrasound, and eddy current images, and (4) x ray, ultrasound, and shearography images.

AB - This paper describes a computational method using tensor math for higher order singular value decomposition (HOSVD) of registered images. Tensor decomposition is a rigorous way to expose structure embedded in multidimensional datasets. Given a dataset of registered 2-D images, the dataset is represented in tensor format and HOSVD of the tensor is computed to obtain a set of 2-D basis images. The basis images constitute a linear decomposition of the original dataset. HOSVD is data-driven and does not require the user to select parameters or assign thresholds. A specific application uses the basis images for pixel-level fusion of registered images into a single image for visualization. The fusion is optimized with respect to a measure of mean squared error. HOSVD and image fusion are illustrated empirically with four real datasets: (1) visible and infrared data of a natural scene, (2) MRI and x ray CT brain images, and in nondestructive testing (3) x ray, ultrasound, and eddy current images, and (4) x ray, ultrasound, and shearography images.

UR - http://www.scopus.com/inward/record.url?scp=80055090464&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80055090464&partnerID=8YFLogxK

U2 - 10.1117/1.3563592

DO - 10.1117/1.3563592

M3 - Article

VL - 20

JO - Journal of Electronic Imaging

JF - Journal of Electronic Imaging

SN - 1017-9909

IS - 1

M1 - 013023

ER -