Competition Code And LBP Palm Vein Feature-Level Fusion Using Canonical Correlation Analysis

Biometrics is a discipline in computer science that uses biometrics to identify people and control access. With the development of technology, a variety of biometric recognition technologies are widely used, such as palm vein recognition technology. However, there are deficiencies in the feature extraction of the palm vein in one way. It is difficult to classify and identify. Therefore, the palm vein features are extracted in two different ways to obtain two different feature sets. And the two feature sets have complementary characteristics when expressing the palm vein features. Then, to improve the classification effect, we used the Canonical Correlation Analysis (CCA) to fuse the two feature sets. The palm vein features are extracted using a Competition Code and a Local Binary Pattern (LBP) to obtain two different palm vein features sets in this paper. The Competition Code uses the local orientation information of the image to extract the palm vein feature, and the LBP utilizes the local texture feature of the image to extract the palm vein feature. These two features can achieve complementarity. CCA is a feature-level fusion technique. CCA projects two feature sets into the same spatial domain through linear transformation, and achieves effective feature fusion in the same spatial domain. A good classification effect can be achieved by the two feature sets of the CCA fusion palm vein. Our experiments are carried out in a public database of Hong Kong Polytechnic University. Compared with the single palm vein competition code feature or LBP feature, the palm vein feature after CCA fusion shortens the classification time in some degree. And the recognition rate is increased to 98.14% when the ratio of the training sample to the test sample is 9:3.

[1]  Matti Pietikäinen,et al.  Face Description with Local Binary Patterns: Application to Face Recognition , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Jie Chen,et al.  Hierarchical particle fi lter tracking algorithm based on multi-feature fusion , 2015 .

[3]  David Zhang,et al.  Competitive coding scheme for palmprint verification , 2004, ICPR 2004.

[4]  Ma Xin,et al.  Palm vein recognition method based on fusion of local Gabor histograms , 2017 .

[5]  Ammad Ali,et al.  Face Recognition with Local Binary Patterns , 2012 .

[6]  Saman A. Zonouz,et al.  Identification Using Encrypted Biometrics , 2013, CAIP.

[7]  Shengcai Liao,et al.  Learning Multi-scale Block Local Binary Patterns for Face Recognition , 2007, ICB.

[8]  Vince D. Calhoun,et al.  Canonical Correlation Analysis for Data Fusion and Group Inferences , 2010, IEEE Signal Processing Magazine.

[9]  Jianpei Zhang,et al.  Uncertain canonical correlation analysis for multi-view feature extraction from uncertain data streams , 2015, Neurocomputing.

[10]  strong,et al.  Hierarchical Particle Filter Tracking AlgorithmBased on Multi-feature Fusion , 2016 .

[11]  Dongmei Sun,et al.  A research on palm vein recognition , 2016, 2016 IEEE 13th International Conference on Signal Processing (ICSP).

[12]  Matti Pietikäinen,et al.  A comparative study of texture measures with classification based on featured distributions , 1996, Pattern Recognit..

[13]  Mohamed Abdel-Mottaleb,et al.  Discriminant Correlation Analysis: Real-Time Feature Level Fusion for Multimodal Biometric Recognition , 2016, IEEE Transactions on Information Forensics and Security.

[14]  Guangfeng Lin,et al.  Heterogeneous feature structure fusion for classification , 2016, Pattern Recognit..

[15]  Anil K. Jain,et al.  Matching Composite Sketches to Face Photos: A Component-Based Approach , 2013, IEEE Transactions on Information Forensics and Security.