Improving SIFT-based descriptors stability to rotations

Risultato della ricerca: Conference contribution

18 Citazioni (Scopus)

Abstract

Image descriptors are widely adopted structures to match image features. SIFT-based descriptors are collections of gradient orientation histograms computed on different feature regions, commonly divided by using a regular Cartesian grid or a log-polar grid. In order to achieve rotation invariance, feature patches have to be generally rotated in the direction of the dominant gradient orientation. In this paper we present a modification of the GLOH descriptor, a SIFT-based descriptor based on a log-polar grid, which avoids to rotate the feature patch before computing the descriptor since predefined discrete orientations can be easily derived by shifting the descriptor vector. The proposed descriptors, called sGLOH and sGLOH+, have been compared with the SIFT descriptor on the Oxford image dataset, with good results which point out its robustness and stability
Lingua originaleEnglish
Titolo della pubblicazione ospitePattern Recognition, International Conference on
Pagine3460-3463
Numero di pagine4
Stato di pubblicazionePublished - 2010

Serie di pubblicazioni

NomeINTERNATIONAL CONFERENCE ON PATTERN RECOGNITION

Fingerprint

Invariance

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cita questo

Bellavia, F., Tegolo, D., & Trucco, E. (2010). Improving SIFT-based descriptors stability to rotations. In Pattern Recognition, International Conference on (pagg. 3460-3463). (INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION).

Improving SIFT-based descriptors stability to rotations. / Bellavia, Fabio; Tegolo, Domenico; Trucco, Emanuele.

Pattern Recognition, International Conference on. 2010. pag. 3460-3463 (INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION).

Risultato della ricerca: Conference contribution

Bellavia, F, Tegolo, D & Trucco, E 2010, Improving SIFT-based descriptors stability to rotations. in Pattern Recognition, International Conference on. INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION, pagg. 3460-3463.
Bellavia F, Tegolo D, Trucco E. Improving SIFT-based descriptors stability to rotations. In Pattern Recognition, International Conference on. 2010. pag. 3460-3463. (INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION).
Bellavia, Fabio ; Tegolo, Domenico ; Trucco, Emanuele. / Improving SIFT-based descriptors stability to rotations. Pattern Recognition, International Conference on. 2010. pagg. 3460-3463 (INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION).
@inproceedings{97e67b50184e4a8b845c31113a37d480,
title = "Improving SIFT-based descriptors stability to rotations",
abstract = "Image descriptors are widely adopted structures to match image features. SIFT-based descriptors are collections of gradient orientation histograms computed on different feature regions, commonly divided by using a regular Cartesian grid or a log-polar grid. In order to achieve rotation invariance, feature patches have to be generally rotated in the direction of the dominant gradient orientation. In this paper we present a modification of the GLOH descriptor, a SIFT-based descriptor based on a log-polar grid, which avoids to rotate the feature patch before computing the descriptor since predefined discrete orientations can be easily derived by shifting the descriptor vector. The proposed descriptors, called sGLOH and sGLOH+, have been compared with the SIFT descriptor on the Oxford image dataset, with good results which point out its robustness and stability",
author = "Fabio Bellavia and Domenico Tegolo and Emanuele Trucco",
year = "2010",
language = "English",
isbn = "978-1-4244-7541-4",
series = "INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION",
pages = "3460--3463",
booktitle = "Pattern Recognition, International Conference on",

}

TY - GEN

T1 - Improving SIFT-based descriptors stability to rotations

AU - Bellavia, Fabio

AU - Tegolo, Domenico

AU - Trucco, Emanuele

PY - 2010

Y1 - 2010

N2 - Image descriptors are widely adopted structures to match image features. SIFT-based descriptors are collections of gradient orientation histograms computed on different feature regions, commonly divided by using a regular Cartesian grid or a log-polar grid. In order to achieve rotation invariance, feature patches have to be generally rotated in the direction of the dominant gradient orientation. In this paper we present a modification of the GLOH descriptor, a SIFT-based descriptor based on a log-polar grid, which avoids to rotate the feature patch before computing the descriptor since predefined discrete orientations can be easily derived by shifting the descriptor vector. The proposed descriptors, called sGLOH and sGLOH+, have been compared with the SIFT descriptor on the Oxford image dataset, with good results which point out its robustness and stability

AB - Image descriptors are widely adopted structures to match image features. SIFT-based descriptors are collections of gradient orientation histograms computed on different feature regions, commonly divided by using a regular Cartesian grid or a log-polar grid. In order to achieve rotation invariance, feature patches have to be generally rotated in the direction of the dominant gradient orientation. In this paper we present a modification of the GLOH descriptor, a SIFT-based descriptor based on a log-polar grid, which avoids to rotate the feature patch before computing the descriptor since predefined discrete orientations can be easily derived by shifting the descriptor vector. The proposed descriptors, called sGLOH and sGLOH+, have been compared with the SIFT descriptor on the Oxford image dataset, with good results which point out its robustness and stability

UR - http://hdl.handle.net/10447/226245

M3 - Conference contribution

SN - 978-1-4244-7541-4

T3 - INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION

SP - 3460

EP - 3463

BT - Pattern Recognition, International Conference on

ER -