Liveness Detection Of Dorsal Hand Vein Recognition Based On Multi-Spectrum Technology

With the rapid development of society, science and technology has been used in every corner of people’s life, smart homes, smart phones, intelligent traffic, etc, science and technology is gradually changing the routine of life to let people known. As identity authentication, people more thinked is signature, keys, etc., but as a biometric authentication method is puting forward, people with the idea of identity in a little bit of change, gradually from the key to fingerprint. As a kind of has the characteristics of each person is born, biometric authentication method has the incomparable superiority to traditional authentication methods, don’t have to worry about forgetting, do not need to sign and as long as press fingers can identify.As biological characteristics have been proposed in the 1980 s, its in just 30 years of rapid development, at the same time also received attention and support of social from all walks of life. As has been relatively mature technology of fingerprint and face recognition has been widely applied in our life, such as traffic vehicle monitoring, airport traffic monitoring, etc. In recent years, dorsal hand vein recognition has got rapidly development, as other features of hand vein did, both in recognition algorithm and application has toke into a curved upward trend.This paper proposes a method which combined dorsal hand vein feature extraction and double infrared band detection, the feature distribution of the vein is differences in different infrared absorption spectrum in depth analysis of selected 850 nm and 1100 nm as a band with the biggest distinguish effect to collect samples after collection of the image feature extraction, the amount of the vein characteristics among the different infrared wavelengths as a distinguishing characteristic opponent back in live testing. Verified by the experimental recognition rate reached 98.1%,reached an ideally liveness detection effectiveness. The mainly innovation works of this paper are as follow:1) Selecting the appropriate range of infrared wavelengths to make analysis as a whole to find the authenticity infrared absorption band of the dorsal hand vein samples, so as to determine the maximum absorption band and minimum absorption band.2) According to the band selected in step 1 to conform to the conditions of the near infrared camera and infrared optical filter. In order to test the feasibility of expermiment, it is need to take the price of camera and infrared optional filters into considered, therefor, the result of the experiment is compared to multiple vendors overall and choose after shooting effect, practicability and feasibility were carried out under ideal conditions.3) Liveness detection for dorsal hand vein recognition is based on the authenticity of the dorsal hand vein samples is different between two different infrared absorption band, cause vein characteristic under the premise of change, in order to be able to best separate among the condition of two kinds of vein area, choose feature points after a large number of experiments by the pixel points of palm-dorsal vein image, as a distinguishing factor can maximum limit and the most effective avoid environmental factors such as external causes the effects on the classification.