The Design And Implementation Of The Finger Vein Recognition Prototype System

With the rapid development of information technology in 21st century, possessing rich and accurate information resources is particularly important. But at the same time, the hidden danger in information security caused by information resources has become more and more obvious. Just rely on traditional identification ways, such as personal certification, password, and other means to ensure the security of information is insufficient. How to identify one's identity conveniently and accurately? How to guarantee information security? More than these problems need to be solved urgently. The identification technologies which rely on inherent biometrics in human body, is developing rapidly, conforming to the trend of the times. Finger Vein is one of the human biometrics, locating in human hands, on the one hand, the images are easy to obtain, on the other hand, it is not easy to forge or destroy the finger veins.Because the finger vein is difficult to replicate and convenient to gather, the finger vein recognition technology, with advanced technology, has vast applied future. Although there are masses of mature products in our life, they have some drawbacks, such as low recognition rate and reaction speed. The key points leading to these problems lie in processing algorithm on finger vein images. So, the core problem in finger vein recognition progressing is to study the processing algorithm intensively and to promote the performance of core algorithms. And algorithm studying needs a corresponding condition and a work platform. With this background, the paper designed and realized the finger vein recognition prototype system. The prototype system provides a foundation platform for the further study of various core algorithms on finger vein recognition technology. So, it has important value for finger vein recognition theory study.The paper designed and realized the finger vein recognition prototype system mainly. The work included the functional and the performance requirement analysis, the system design and realization based on the requirement analysis. Due to the core content of the prototype system were the functional modules which realized the finger vein recognition algorithms, the work of this paper mainly focused on these functional modules. The paper is organized as follows:1. The function of finger vein images grayscale change was studied. The function can translate color images of 24 bits into gray images of 8 bits successfully.2. The function of finger vein images segmentation was realized. Through the segmentation function, the finger image can be extracted from the original image.3. The normalization is carried out to the finger vein image through image's size and grayscale.4. In order to reduce the influence of noise points to the finger veins, and to enlarge the gray-value difference between finger veins and noise points, the enhancement function of finger vein images was studied.5. In order to reduce the time and space loss in image processing, and to further improve the performance of the prototype system, the image binarization function and the image thinning function were studied and realized.6. The feature extraction function was achieved to obtain the bifurcation points and endpoints from the skeletonized image of finger veins.7. The match function which was used to match images of homologous and nonhomologous finger veins was implemented by means of match algorithm with moment invariant features.