TY - GEN
T1 - A boosted cascade of directional local binary patterns for multispectral palmprint recognition
AU - Shen, Linlin
AU - Liu, Bojie
AU - He, Jinwen
PY - 2013
Y1 - 2013
N2 - In this paper, a recently developed local feature descriptor, namely directional local binary patterns (DLBP), was first proposed for palmprint recognition. Compared with local binary patterns (LBP) and directional binary code (DBC), DLBP contains more information on both edge and texture. A cascade structure using AdaBoost algorithm is then used to reduce the feature dimension of DLBP and computational costs of classification. The proposed approach was applied to fuse multispectral palmprint images captured under red, green, blue and near-infrared (NIR) lighting sources for personal identification. Experimental results suggest that the proposed algorithm performs much better than DBC, LBP and PalmCode in identifying palmprint images captured using different illuminations. When fusing the multispectral images, the proposed approach has also been shown to achieve higher accuracy than other methods in literature such as QPCA (Quaternion PCA) and QDWT (Quaternion Discrete Wavelet Transform).
AB - In this paper, a recently developed local feature descriptor, namely directional local binary patterns (DLBP), was first proposed for palmprint recognition. Compared with local binary patterns (LBP) and directional binary code (DBC), DLBP contains more information on both edge and texture. A cascade structure using AdaBoost algorithm is then used to reduce the feature dimension of DLBP and computational costs of classification. The proposed approach was applied to fuse multispectral palmprint images captured under red, green, blue and near-infrared (NIR) lighting sources for personal identification. Experimental results suggest that the proposed algorithm performs much better than DBC, LBP and PalmCode in identifying palmprint images captured using different illuminations. When fusing the multispectral images, the proposed approach has also been shown to achieve higher accuracy than other methods in literature such as QPCA (Quaternion PCA) and QDWT (Quaternion Discrete Wavelet Transform).
KW - Directional local binary patterns
KW - Fusion
KW - Palmprint recognition
UR - http://www.scopus.com/inward/record.url?scp=84893064062&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-02961-0_29
DO - 10.1007/978-3-319-02961-0_29
M3 - Conference contribution
AN - SCOPUS:84893064062
SN - 9783319029603
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 233
EP - 240
BT - Biometric Recognition - 8th Chinese Conference, CCBR 2013, Proceedings
T2 - 2012 International Conference on Service-Oriented Computing, ICSOC 2012
Y2 - 16 November 2013 through 17 November 2013
ER -