Optimizing the selection of a biometric template from a sequence

The problem of choosing the best biometric template from a group constructed based on a sequence of registered images is addressed. A method based on the analysis of the matrix of distances of a group of templates is proposed. A comparison with the standard approaches using the quality indices of th...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Journal of computer & systems sciences international Ročník 54; číslo 3; s. 399 - 405
Hlavní autoři: Gneushev, A. N., Kovkov, D. V., Matveev, I. A., Novik, V. P.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Moscow Pleiades Publishing 01.05.2015
Springer Nature B.V
Témata:
Analysis
Biometrics
Computer science
Control
Engineering
Image processing systems
Mechatronics
Pattern Recognition and Image Processing
Robotics
Studies
Quality Index
Iris Image
Biometric Template
Iris Recognition
System Science International
ISSN:1064-2307, 1555-6530
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The problem of choosing the best biometric template from a group constructed based on a sequence of registered images is addressed. A method based on the analysis of the matrix of distances of a group of templates is proposed. A comparison with the standard approaches using the quality indices of the initial image is performed. It is specified that the quality indices of the image are developed for the problem of rejecting poor images and little suit the problem of choosing the best image. The computer experiments were conducted based on several databases available in open source with more than 70000 images. The tests have shown that the proposed method provides a slightly better quality of the chosen templates. Note that the method does not require the development of additional quality measures and employs the calculations of the distance available.
Bibliografie:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ISSN:1064-2307
1555-6530
DOI:10.1134/S1064230715030077