Liveness Detection Of Hand Vein Recognition Based On Transform Domain

Nowadays information technologies develop so rapidly that the importance of information security becomes more and more obvious. The problem of how to effectively protect the personal information has been paid considerable attention. The traditional identity authentication modes including keys, magnetic cards and commands, has many shortcomings because they are too simple, easy to be lost and forgotten, so they are difficult to meet the needs of the current information society. Biometric technology is based on physiological or behavioral characteristics, such as fingerprint, face, iris, voice, vein etc. to confirm personal identity. Compared with traditional personal recognition methods, biometrics is unique, universal and permanent. Governments and organizations around the world paid huge manpower and material resources in the biometric field. Nowadays, biometric technology has been widely used in our life and serves as identification system in many financial institutions and public places.Among all biometrics, dorsal hand vein (DHV) recognition has gained considerable attentions because its low cost in device, non-contact, immune to blots and scratches, etc. It’s originated in the late1980s and has entered a stage of rapid development. Its image-forming principle is based on the different absorptivity to the infrared light of blood and surrounding tissues. And then the feature extraction and verification is carried out through image processing and pattern recognition. The DHV recognition system has been used in many airports, banks and hospitals. However, DHV recognition is vulnerable to spoofing fake samples such as printed DHV images or rubber gloves drawn with DHV structures from individuals.In this paper, we introduce the image-forming principle and associated image capture device and discuss the image preprocessing methods. Against the issue that DHV recognition system might be attacked by the forged sample, this paper proposed a novel method for liveness detection of dorsal hand vein. The major innovations of our work can be summarized as follows:1. The dorsal hand vein database was enriched in our work including printed DHV images and wearing rubber gloves. And then the effectiveness of fake images was investigated to make sure our database has fake samples which could successfully cheated recognition system. 2. Aiming at the problem of recognition system might be attacked by forging sample, we proposed a novel liveness detection method of dorsal hand vein. Firstly, we discuss the difference of real and fake images because of image acquisition, and based on the method of PCA, test samples are projected to produce one-dimensional vectors which include the noise of DHV images. Then, an AR model established for these vector by estimating their power spectrums. At last, we used the power spectrum to distinguish the fake from the real.