David Zhang (Hong Kong Polytechnic University, Hong Kong), Fengxi Song (New Star Research Institute Of Applied Technology, China), Yong Xu (Harbin Institute of Technology, China) and Zhizhen Liang (Shanghai Jiao Tong University, China)
Copyright: © 2009
With the title “Advanced Pattern Recognition Technologies with Applications to Biometrics” this book mainly focuses on two kinds of advanced biometric recognition technologies, biometric discrimination techniques and multi-biometrics. Biometric discrimination techniques are presented in Parts I and II, while multi-biometrics is described in Part III. While the methods and algorithms described in Parts I and II are very suitable for biometrics as they take into account characteristics of biometric applications such as high dimensionality and small sample size, Part III mainly introduces three kinds of biometric fusion techniques that respectively fuse biometric information at the feature level, matching score level and decision level as well as their applications cases. This chapter summarizes the book from a holistic viewpoint. Section 16.1 summarizes the contents of the book and indicates the relationship between different chapters in each part. Section 16.2 reveals that how the methods and algorithms described in different parts can be applied to different data forms of biometric traits. Section 16.3 provides comments on the development of multi-biometrics.
Biometric data has three typical representation forms: the vector form, the two-dimensional image form and the 3D matrix form. This book also has the potential to provide us with three types of biometric discriminant methods applicable to biometric traits with different representation forms. The methods presented in Part I are suited to biometric data in the vector form. This type of methods is referred to as one-dimensional biometric method. Part II provides a type of methods that is applicable to the two-dimensional image form and is referred to as two-dimensional biometric method. These methods include two-dimensional LPP, two-dimensional PCA and LDA as well as GLRAM, and so on. From Part II, we know that it is also possible for the tensor-based methods to provide us with a tool to directly analyze and exploit biometric data in the 3D matrix form. This novel type of analysis techniques for 3D biometric data can be referred to as three-dimensional biometric method.