Jet-Based Local Image Descriptors

Anders Boesen Lindbo Larsen; Sune Darkner; Anders Lindbjerg Dahl; Kim Steenstrup Pedersen

doi:10.1007/978-3-642-33712-3_46

Jet-Based Local Image Descriptors

Anders Boesen Lindbo Larsen
, Sune Darkner
, Anders Lindbjerg Dahl
, Kim Steenstrup Pedersen

University of Copenhagen

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

Abstract

We present a general novel image descriptor based on higherorder differential geometry and investigate the effect of common descriptor choices. Our investigation is twofold in that we develop a jet-based descriptor and perform a comparative evaluation with current state-of-the-art descriptors on the recently released DTU Robot dataset. We demonstrate how the use of higher-order image structures enables us to reduce the descriptor dimensionality while still achieving very good performance. The descriptors are tested in a variety of scenarios including large changes in scale, viewing angle and lighting. We show that the proposed jet-based descriptor is superior to state-of-the-art for DoG interest points and show competitive performance for the other tested interest points.

Original language	English
Title of host publication	Computer Vision – ECCV 2012 : Workshops and Demonstrations, Part III
Publisher	Springer
Publication date	2012
Pages	638-650
ISBN (Print)	978-3-642-33711-6
ISBN (Electronic)	978-3-642-33712-3
DOIs	https://doi.org/10.1007/978-3-642-33712-3_46
Publication status	Published - 2012
Event	12th European Conference on Computer Vision (ECCV 2012) - Florence, Italy Duration: 7 Oct 2012 → 13 Oct 2012 http://eccv2012.unifi.it/

Conference

Conference	12th European Conference on Computer Vision (ECCV 2012)
Country/Territory	Italy
City	Florence
Period	07/10/2012 → 13/10/2012
Internet address	http://eccv2012.unifi.it/

Series	Lecture Notes in Computer Science
Volume	7584
ISSN	0302-9743

Access to Document

10.1007/978-3-642-33712-3_46

OpenUrl availability

Full text

Cite this

@inproceedings{388ea09d8af44d00ab25952cbd0ec3a1,

title = "Jet-Based Local Image Descriptors",

abstract = "We present a general novel image descriptor based on higherorder differential geometry and investigate the effect of common descriptor choices. Our investigation is twofold in that we develop a jet-based descriptor and perform a comparative evaluation with current state-of-the-art descriptors on the recently released DTU Robot dataset. We demonstrate how the use of higher-order image structures enables us to reduce the descriptor dimensionality while still achieving very good performance. The descriptors are tested in a variety of scenarios including large changes in scale, viewing angle and lighting. We show that the proposed jet-based descriptor is superior to state-of-the-art for DoG interest points and show competitive performance for the other tested interest points. ",

author = "Larsen, \{Anders Boesen Lindbo\} and Sune Darkner and Dahl, \{Anders Lindbjerg\} and Pedersen, \{Kim Steenstrup\}",

year = "2012",

doi = "10.1007/978-3-642-33712-3\_46",

language = "English",

isbn = "978-3-642-33711-6",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "638--650",

booktitle = "Computer Vision – ECCV 2012",

note = "12th European Conference on Computer Vision (ECCV 2012) ; Conference date: 07-10-2012 Through 13-10-2012",

url = "http://eccv2012.unifi.it/",

}

Jet-Based Local Image Descriptors. / Larsen, Anders Boesen Lindbo; Darkner, Sune; Dahl, Anders Lindbjerg et al.
Computer Vision – ECCV 2012: Workshops and Demonstrations, Part III. Springer, 2012. p. 638-650 (Lecture Notes in Computer Science, Vol. 7584).

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

TY - GEN

T1 - Jet-Based Local Image Descriptors

AU - Larsen, Anders Boesen Lindbo

AU - Darkner, Sune

AU - Dahl, Anders Lindbjerg

AU - Pedersen, Kim Steenstrup

PY - 2012

Y1 - 2012

N2 - We present a general novel image descriptor based on higherorder differential geometry and investigate the effect of common descriptor choices. Our investigation is twofold in that we develop a jet-based descriptor and perform a comparative evaluation with current state-of-the-art descriptors on the recently released DTU Robot dataset. We demonstrate how the use of higher-order image structures enables us to reduce the descriptor dimensionality while still achieving very good performance. The descriptors are tested in a variety of scenarios including large changes in scale, viewing angle and lighting. We show that the proposed jet-based descriptor is superior to state-of-the-art for DoG interest points and show competitive performance for the other tested interest points.

AB - We present a general novel image descriptor based on higherorder differential geometry and investigate the effect of common descriptor choices. Our investigation is twofold in that we develop a jet-based descriptor and perform a comparative evaluation with current state-of-the-art descriptors on the recently released DTU Robot dataset. We demonstrate how the use of higher-order image structures enables us to reduce the descriptor dimensionality while still achieving very good performance. The descriptors are tested in a variety of scenarios including large changes in scale, viewing angle and lighting. We show that the proposed jet-based descriptor is superior to state-of-the-art for DoG interest points and show competitive performance for the other tested interest points.

U2 - 10.1007/978-3-642-33712-3_46

DO - 10.1007/978-3-642-33712-3_46

M3 - Article in proceedings

SN - 978-3-642-33711-6

T3 - Lecture Notes in Computer Science

SP - 638

EP - 650

BT - Computer Vision – ECCV 2012

PB - Springer

T2 - 12th European Conference on Computer Vision (ECCV 2012)

Y2 - 7 October 2012 through 13 October 2012

ER -

Jet-Based Local Image Descriptors

Abstract

Conference

Access to Document

OpenUrl availability

Fingerprint

Cite this