Research Of Finger Vein Recognition Technology And Its Implementation

Biometric identification technology is the technology to implement personal identification by using the physiological characteristics of human body. With the development of the society and progress of technology, its application becoming much more widely, such as in airports, banks, military base, R&D centers may be equipped with such biometric recognitions such as fingerprint recognition, face recognition, iris recognition, vein recognition, etc. Among them, finger vein recognition technology is outstanding to be one of the most promising Biometric identification technologies in recent years with its advantages such as high security and accurate, easy to use, fast recognition. Research of finger vein recognition technology has a high application value.Surrounding several key problems on finger vein recognition technology, this paper was carried out with more in-depth analysis and research by focus on the design of finger vein image acquisition device, vein feature extraction and matching algorithms and the application of finger vein authentication system. First of all, for the finger vein imaging methods, the vein imaging results were compared through experiments by using NIR(near-infrared) LED array and NIR laser as the light source. Secondly, based on the study and analysis of several common vein recognition algorithm domestic and abroad, it put forward a vein extraction algorithm by using four-directional valley-shaped gray search, improved the traditional vein matching based on template matching method and compared several extraction algorithms. Finally, a finger vein identification system based on PC was designed to verify and analyze the recognition results.Results of the research show that the method to imaging finger vein proposed in this paper which using NIR laser as the light source can achieve better quality vein image compared with traditional method using NIR LED array. Through real testing on our laboratory members, it was found that the finger vein recognition algorithms present in this paper has good extraction effect, high identifying accuracy, and it can meet the demands of real-time in vein identfication.