Intensity-Based Congealing for Unsupervised Joint Image Alignment

Markus Storer, Martin Urschler, Horst Bischof

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

Abstract

We present an approach for unsupervised alignment of an ensemble of images called congealing. Our algorithm is based on image registration using the mutual information measure as a cost function. The cost function is optimized by a standard gradient descent method in a multiresolution scheme. As opposed to other congealing methods, which use the SSD measure, the mutual information measure is better suited as a similarity measure for registering images since no prior assumptions on the relation of intensities between images are required. We present alignment results on the MNIST handwritten digit database and on facial images obtained from the CVL database.
Originalspracheenglisch
Titel20th International IEEE Conference on Pattern Recognition (ICPR)
Herausgeber (Verlag)Institute of Electrical and Electronics Engineers
Seiten1473-1476
ISBN (Print)978-1-4244-7542-1
DOIs
PublikationsstatusVeröffentlicht - 2010

Fingerprint

Cost functions
Image registration

Fields of Expertise

  • Information, Communication & Computing

Dies zitieren

Storer, M., Urschler, M., & Bischof, H. (2010). Intensity-Based Congealing for Unsupervised Joint Image Alignment. in 20th International IEEE Conference on Pattern Recognition (ICPR) (S. 1473-1476). Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICPR.2010.364

Intensity-Based Congealing for Unsupervised Joint Image Alignment. / Storer, Markus; Urschler, Martin; Bischof, Horst.

20th International IEEE Conference on Pattern Recognition (ICPR). Institute of Electrical and Electronics Engineers, 2010. S. 1473-1476.

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

Storer, M, Urschler, M & Bischof, H 2010, Intensity-Based Congealing for Unsupervised Joint Image Alignment. in 20th International IEEE Conference on Pattern Recognition (ICPR). Institute of Electrical and Electronics Engineers, S. 1473-1476. https://doi.org/10.1109/ICPR.2010.364
Storer M, Urschler M, Bischof H. Intensity-Based Congealing for Unsupervised Joint Image Alignment. in 20th International IEEE Conference on Pattern Recognition (ICPR). Institute of Electrical and Electronics Engineers. 2010. S. 1473-1476 https://doi.org/10.1109/ICPR.2010.364
Storer, Markus ; Urschler, Martin ; Bischof, Horst. / Intensity-Based Congealing for Unsupervised Joint Image Alignment. 20th International IEEE Conference on Pattern Recognition (ICPR). Institute of Electrical and Electronics Engineers, 2010. S. 1473-1476
@inproceedings{c3fcbfd369db45d2acf92c9cbccc1fa3,
title = "Intensity-Based Congealing for Unsupervised Joint Image Alignment",
abstract = "We present an approach for unsupervised alignment of an ensemble of images called congealing. Our algorithm is based on image registration using the mutual information measure as a cost function. The cost function is optimized by a standard gradient descent method in a multiresolution scheme. As opposed to other congealing methods, which use the SSD measure, the mutual information measure is better suited as a similarity measure for registering images since no prior assumptions on the relation of intensities between images are required. We present alignment results on the MNIST handwritten digit database and on facial images obtained from the CVL database.",
author = "Markus Storer and Martin Urschler and Horst Bischof",
year = "2010",
doi = "10.1109/ICPR.2010.364",
language = "English",
isbn = "978-1-4244-7542-1",
pages = "1473--1476",
booktitle = "20th International IEEE Conference on Pattern Recognition (ICPR)",
publisher = "Institute of Electrical and Electronics Engineers",
address = "United States",

}

TY - GEN

T1 - Intensity-Based Congealing for Unsupervised Joint Image Alignment

AU - Storer, Markus

AU - Urschler, Martin

AU - Bischof, Horst

PY - 2010

Y1 - 2010

N2 - We present an approach for unsupervised alignment of an ensemble of images called congealing. Our algorithm is based on image registration using the mutual information measure as a cost function. The cost function is optimized by a standard gradient descent method in a multiresolution scheme. As opposed to other congealing methods, which use the SSD measure, the mutual information measure is better suited as a similarity measure for registering images since no prior assumptions on the relation of intensities between images are required. We present alignment results on the MNIST handwritten digit database and on facial images obtained from the CVL database.

AB - We present an approach for unsupervised alignment of an ensemble of images called congealing. Our algorithm is based on image registration using the mutual information measure as a cost function. The cost function is optimized by a standard gradient descent method in a multiresolution scheme. As opposed to other congealing methods, which use the SSD measure, the mutual information measure is better suited as a similarity measure for registering images since no prior assumptions on the relation of intensities between images are required. We present alignment results on the MNIST handwritten digit database and on facial images obtained from the CVL database.

U2 - 10.1109/ICPR.2010.364

DO - 10.1109/ICPR.2010.364

M3 - Conference contribution

SN - 978-1-4244-7542-1

SP - 1473

EP - 1476

BT - 20th International IEEE Conference on Pattern Recognition (ICPR)

PB - Institute of Electrical and Electronics Engineers

ER -