Category-preserving binary feature learning and binary codebook learning for finger vein recognition
- PDF / 1,435,711 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 47 Downloads / 192 Views
ORIGINAL ARTICLE
Category‑preserving binary feature learning and binary codebook learning for finger vein recognition Haiying Liu1 · Gongping Yang2 · Yilong Yin2 Received: 21 January 2020 / Accepted: 12 May 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Local binary feature learning has attracted a lot of researches in image recognition due to its vital effectiveness. Generally, in the traditional local feature learning methods, a projection is learned to map the patches of image into binary features and then a codebook is generated by clustering the binary features with K-means clustering. However, these local feature learning methods, such as compact binary face descriptor and discriminative binary descriptor, ignore the category specific distributions of the original features during the feature learning process and use the real-valued clustering approach to generate the codebook, the discriminant of feature is degraded and the merits of binary feature are lost. To tack these problems, in this paper, we propose a novel category-preserving binary feature learning and binary codebook leaning (CPBFL-BCL) method for finger vein recognition. In CPBFL-BCL, the discrimination of learned binary features is generated by criteria of fisher discriminant analysis and category manifold preserving regularity during the feature learning process, and a novel binary clustering method based on K-means clustering is designed to generate binary codebook. Experimental results on recognition and retrieval tasks using two public finger vein databases are presented and demonstrate the effectiveness and efficiency of the proposed method over the state-of-the-art finger vein methods and a finger vein retrieval method. Keywords Finger vein recognition · Manifold structure · Category preservation · Binary feature learning · Codebook learning
1 Introduction Biometrics is the technology for authenticating human identity on the basis of intrinsic physical features and behaviour traits, such as finger print, palm print [1], iris, face [2], finger vein [3], palm vein [1], gait, Electrocardiogram (ECG) and so on. Among these technologies, finger vein recognition technology has several advantages over the other methods: (1) Security. It is difficult to forge and copy the finger vein image to attack a recognition system because complicate finger vein patterns lie beneath the surface of the human body and only live human can be captured the finger vein * Gongping Yang [email protected] Haiying Liu [email protected] 1
Department of Computer Engineering, Changji University, Changji 831100, China
School of Software, Shandong University, Jinan 250101, China
2
image [1]. (2) The small capture device. The finger vein capture device is miniature and convenient to use in different practical scenarios. (3) It is easy to be accepted by people. When capturing a finger vein image, there is no discomfort to the user, and the finger vein image does not reveal personal privacy, such as health condition or disease. For th
Data Loading...
