The research of double-biometric identification technology based on finger geometry & palm print

Common system of single biometric identification is based on single technology, such as fingerprints identification, face identification and iris identification. They have unicity and limitations. So multi-biometric identification technology is valued extensively and becomes the focus of studies. This article, which bases on the features of singular creasure, puts forward multi-biometric identification technology based on finger shape and palm print. It employs different methods of feature collection in view of different areas and designs a hierarchical matcher from coarse to fine so that a multi-biometric identification system is realized with high recognition accuracy and good extendibility. The experimental results demonstrate that this multi-biometric identification technology has not only greatly increased the efficiency of identifying but also ensured the accuracy and security of identification.

I. Introduction

In the highly informatized developed commercial society, there is a great many need of authentication for human beings every day, such as assess control, checks on attendance, management of credit exchanging and so on. As products of fingerprint identification technology behave well in their levels of security and convenience, they have already included about 50 percent share of market [4]. However, with the extensive application of XL-URK, there are more and more malicious decryption and defective application. Among them, the “fingerprint cover” which is called “assistant card reader” or “finger membrane” has been widely used. After referring to fingerprint samples, silica gel is dealt with simple procedure of rubbing and mould reversing. Then the fingerprint cover can easily escapes from the “light eyes” of the card reader with a rate as high as 90 percent. What's more, as a following step, some lawbreaking operations even plan to develop “fingerprint finger” to substitute existing products. In the face of this, fingerprint card readers have already completely lost the goals which they should reach.