A Filtering Method for SIFT based Palm Vein Recognition

Kaveen Perera; Fouad Khelifi; Ammar Belatreche

doi:10.1109/dicta56598.2022.10034589

A Filtering Method for SIFT based Palm Vein Recognition

Kaveen Perera, Fouad Khelifi, Ammar Belatreche

Computer Information Sciences Department

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

41 Downloads (Pure)

Abstract

A key issue with palm vein images is that slight movements of fingers and the thumb or changes in the hand pose can stretch the skin in different areas and alter the vein patterns. This can produce palm vein images with an infinite number of variations for a given subject. This paper presents a novel filtering method for SIFT-based feature matching referred to as the Mean and Median Distance (MMD) Filter, which checks the difference of keypoint coordinates and calculates the mean and the median in each direction in order to filter out the incorrect matches. Experiments conducted on the 850nm subset of the CASIA dataset show that the proposed MMD filter can maintain correct points and reduce false positives that were detected by other filtering methods. Comparison against existing SIFT-based palm vein recognition systems demonstrates that the proposed MMD filter produces excellent performance recording lower Equal Error Rate (EER) values.

Original language	English
Title of host publication	2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA)
Place of Publication	Piscataway, NJ
Publisher	IEEE
Pages	1-7
Number of pages	7
ISBN (Electronic)	9781665456425
ISBN (Print)	9781665456432
DOIs	https://doi.org/10.1109/dicta56598.2022.10034589
Publication status	Published - 30 Nov 2022
Event	International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2022 - Rydges World Square, Sydney, Australia Duration: 30 Nov 2022 → 2 Dec 2022 http://dicta2022.dictaconference.org/index.html

Publication series

Name	Proceedings of the Digital Image Computing: Technqiues and Applications (DICTA)
Publisher	IEEE

Conference

Conference	International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2022
Country/Territory	Australia
City	Sydney
Period	30/11/22 → 2/12/22
Internet address	http://dicta2022.dictaconference.org/index.html

Keywords

MMD filter
SIFT descriptor matching
SIFT filtering
Palm Vein recognition

Access to Document

10.1109/dicta56598.2022.10034589

A Filtering Method for SIFT based Palm Vein Recognition-K Perera F Khelifi A Belatreche-Final_Camera_readyAccepted author manuscript, 1.19 MB

Cite this

@inproceedings{a4c34f3646424d2b9c41caa2d37185f6,

title = "A Filtering Method for SIFT based Palm Vein Recognition",

abstract = "A key issue with palm vein images is that slight movements of fingers and the thumb or changes in the hand pose can stretch the skin in different areas and alter the vein patterns. This can produce palm vein images with an infinite number of variations for a given subject. This paper presents a novel filtering method for SIFT-based feature matching referred to as the Mean and Median Distance (MMD) Filter, which checks the difference of keypoint coordinates and calculates the mean and the median in each direction in order to filter out the incorrect matches. Experiments conducted on the 850nm subset of the CASIA dataset show that the proposed MMD filter can maintain correct points and reduce false positives that were detected by other filtering methods. Comparison against existing SIFT-based palm vein recognition systems demonstrates that the proposed MMD filter produces excellent performance recording lower Equal Error Rate (EER) values.",

keywords = "MMD filter, SIFT descriptor matching, SIFT filtering, Palm Vein recognition",

author = "Kaveen Perera and Fouad Khelifi and Ammar Belatreche",

year = "2022",

month = nov,

day = "30",

doi = "10.1109/dicta56598.2022.10034589",

language = "English",

isbn = "9781665456432",

series = "Proceedings of the Digital Image Computing: Technqiues and Applications (DICTA)",

publisher = "IEEE",

pages = "1--7",

booktitle = "2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA)",

address = "United States",

note = "International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2022 ; Conference date: 30-11-2022 Through 02-12-2022",

url = "http://dicta2022.dictaconference.org/index.html",

}

Perera, K, Khelifi, F & Belatreche, A 2022, A Filtering Method for SIFT based Palm Vein Recognition. in 2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA). Proceedings of the Digital Image Computing: Technqiues and Applications (DICTA), IEEE, Piscataway, NJ, pp. 1-7, International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2022, Sydney, New South Wales, Australia, 30/11/22. https://doi.org/10.1109/dicta56598.2022.10034589

A Filtering Method for SIFT based Palm Vein Recognition. / Perera, Kaveen; Khelifi, Fouad ; Belatreche, Ammar.
2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA). Piscataway, NJ: IEEE, 2022. p. 1-7 (Proceedings of the Digital Image Computing: Technqiues and Applications (DICTA)).

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

TY - GEN

T1 - A Filtering Method for SIFT based Palm Vein Recognition

AU - Perera, Kaveen

AU - Khelifi, Fouad

AU - Belatreche, Ammar

PY - 2022/11/30

Y1 - 2022/11/30

N2 - A key issue with palm vein images is that slight movements of fingers and the thumb or changes in the hand pose can stretch the skin in different areas and alter the vein patterns. This can produce palm vein images with an infinite number of variations for a given subject. This paper presents a novel filtering method for SIFT-based feature matching referred to as the Mean and Median Distance (MMD) Filter, which checks the difference of keypoint coordinates and calculates the mean and the median in each direction in order to filter out the incorrect matches. Experiments conducted on the 850nm subset of the CASIA dataset show that the proposed MMD filter can maintain correct points and reduce false positives that were detected by other filtering methods. Comparison against existing SIFT-based palm vein recognition systems demonstrates that the proposed MMD filter produces excellent performance recording lower Equal Error Rate (EER) values.

AB - A key issue with palm vein images is that slight movements of fingers and the thumb or changes in the hand pose can stretch the skin in different areas and alter the vein patterns. This can produce palm vein images with an infinite number of variations for a given subject. This paper presents a novel filtering method for SIFT-based feature matching referred to as the Mean and Median Distance (MMD) Filter, which checks the difference of keypoint coordinates and calculates the mean and the median in each direction in order to filter out the incorrect matches. Experiments conducted on the 850nm subset of the CASIA dataset show that the proposed MMD filter can maintain correct points and reduce false positives that were detected by other filtering methods. Comparison against existing SIFT-based palm vein recognition systems demonstrates that the proposed MMD filter produces excellent performance recording lower Equal Error Rate (EER) values.

KW - MMD filter

KW - SIFT descriptor matching

KW - SIFT filtering

KW - Palm Vein recognition

UR - https://ieeexplore.ieee.org/xpl/conhome/1001512/all-proceedings

UR - http://dicta2022.dictaconference.org/index.html

U2 - 10.1109/dicta56598.2022.10034589

DO - 10.1109/dicta56598.2022.10034589

M3 - Conference contribution

SN - 9781665456432

T3 - Proceedings of the Digital Image Computing: Technqiues and Applications (DICTA)

SP - 1

EP - 7

BT - 2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA)

PB - IEEE

CY - Piscataway, NJ

T2 - International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2022

Y2 - 30 November 2022 through 2 December 2022

ER -

A Filtering Method for SIFT based Palm Vein Recognition

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this