The Study For Recognition Method Of Infrared Vein Imaging Based On Invariant Moment

With the development of the information society, an increasing number of personal identity recognition methods have appeared. Especially, the aspect of personal identity information recognition and privacy protection has become increasingly prominent. The traditional recognition method has the disadvantages of being lost, counterfeited and cracked easily. These defects will bring the great risks of the personal privacy, property safety and the great obstacles of the development of e-commerce. So it is necessary for us to find some identifying methods with a high recognition rates as well as safety and stability to replace the traditional methods. The biological characteristics of the human body are inherent and entrenched. That is to say that the different individuals have different biological characteristics which can open up a new prospect of the applications of biological information.Although the biometric technology has been gradually applied in reality, there are disadvantages of each kind of biological recognition technology, such as the vulnerabilities of biological characteristics. Nowadays vein recognition technology is a new identification method. Comparing with the other identification methods, there are some advantages. Firstly, it is highly stable. For instance, the veins of the human body do not change with the age unless there is a heavy damage. Secondly, it is strongly unique that there are great differences between the different human veins. Besides, it can be identified only with the condition of blood flowing in the vein and the vein distribution in the subcutaneous tissue is not easy to be imitated. In addition, the finger veins are collected easily and the acquisition equipment has the characteristics of low price, small size and convenient popularization. This paper mainly studies the infrared imaging characteristics and recognition algorithm of the finger vein. Finger vein imaging is mainly from the vein imaging principle. It compares the different lighting ways and its wavelength to find the combination of good imaging effects. Image acquisition of high quality is the premise of recognition. Then the images are done some corresponding pretreatment, such as the image target location, size normalization, intensity normalization, image enhancement, image segmentation and so on. We refine and reprocess the extracted image using a better method called the NiBlack dynamic threshold method. Image matching based on invariant moments introduces the NMI (normalized moment of inertia of the image) features. It can reduce the quantity of operation and can guarantee the accuracy of invariant moments, so the recognition speed is greatly improved.