Off-Line Signature Verification Based on Multi-Scale Local Structural Pattern

被引:6
作者
Wen, Jing [1 ,2 ]
Fang, Bin [1 ]
Zhang, Liang [1 ]
Zhu, Yu [1 ]
机构
[1] Chongqing Univ, Coll Comp Sci, Chongqing 400030, Peoples R China
[2] Chengdu Univ, Key Lab Pattern Recognit & Intelligent Informat P, Inst Higher Educ Sichuan Prov, Chengdu, Peoples R China
关键词
Off-line signature verification; local structural pattern; encoding feature; probability distribution function; Chi-square distance; pattern recognition; FISHER DISCRIMINANT METHOD; WRITER IDENTIFICATION; FACE RECOGNITION; INFORMATION; FEATURES; WAVELET; STATE; ART;
D O I
10.1142/S0218001417560109
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The paper proposes a theoretically very simple, and effective multi-scale local structural patterns (MS-LSP) for off-line signature verification. The proposed strategy models the signature by probabilistically counting the distribution of each local structural pattern on multiple scales. This representation correlates with the writing style of an individual and depict the shape and structural information of signatures. Extensive experiments on two challenging databases (MCYT, GPDS960) show that the proposed approach achieves noticeably verification rates and it is expected to provide a powerful discriminative representation of the signature.
引用
收藏
页数:16
相关论文
共 37 条
[1]  
[Anonymous], 2009, INT J PATTERN RECOGN
[2]  
Batista L., 2008, PATTERN RECOGN, P39
[3]   Reducing forgeries in writer-independent off-line signature verification through ensemble of classifiers [J].
Bertolini, D. ;
Oliveira, L. S. ;
Justino, E. ;
Sabourin, R. .
PATTERN RECOGNITION, 2010, 43 (01) :387-396
[4]   Discriminative DCT : An Efficient and Accurate Approach for Off-line Signature Verification [J].
Bharathi, R. K. ;
Shekar, B. H. .
2014 FIFTH INTERNATIONAL CONFERENCE ON SIGNAL AND IMAGE PROCESSING (ICSIP 2014), 2014, :179-184
[5]  
Bharathi RK, 2013, 2013 INTERNATIONAL CONFERENCE ON ADVANCES IN COMPUTING, COMMUNICATIONS AND INFORMATICS (ICACCI), P2063, DOI 10.1109/ICACCI.2013.6637499
[6]   Text-independent writer identification and verification using textural and allographic features [J].
Bulacu, Marius ;
Schomaker, Lambert .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2007, 29 (04) :701-717
[7]  
Cantrell C. D., 2000, Modern Mathematical Methods for Physicists and Engineers
[8]   SUPERVISED REGULARIZATION LOCALITY-PRESERVING PROJECTION METHOD FOR FACE RECOGNITION [J].
Chen, Wen-Sheng ;
Wang, Wei ;
Yang, Jian-Wei ;
Tang, Yuan Yan .
INTERNATIONAL JOURNAL OF WAVELETS MULTIRESOLUTION AND INFORMATION PROCESSING, 2012, 10 (06)
[9]   Kernel machine-based rank-lifting regularized discriminant analysis method for face recognition [J].
Chen, Wen-Sheng ;
Yuen, Pong Chi ;
Xie, Xuehui .
NEUROCOMPUTING, 2011, 74 (17) :2953-2960
[10]   Two-step single parameter regularization Fisher discriminant method for face recognition [J].
Chen, WS ;
Yuen, PC ;
Huang, J ;
Fang, B .
INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE, 2006, 20 (02) :189-207