Vein Based On Sift Feature Extraction Identity Authentication Technology Research

With the development of computer technology,information technology has brought better services to people,especially in identity authentication.As human beings,human beings not only have natural physiological characteristics,but also behavioral characteristics,which can be used as the basis for information technology identification.There are many kinds of common biometrics,such as fingerprints,faces,veins,and so on.As a biological feature that is hidden inside the human body and difficult to modify,the finger vein is highly safe and can be easily obtained.The SIFT algorithm is widely used for image recognition due to its anti-jamming,anti-rotation,anti-rotation and other characteristics.Among all the vein features,the amount of finger vein information is small,and when the resolution of the captured image is low,the interference is more.However,compared with other vein recognition,the finger vein has the advantages of convenient collection,simple equipment and the like.In this paper,by comparing the common image segmentation methods,a variety of methods are combined to find a better method for processing vein images,so that the feature extraction of SIFT can be carried out later.Then,in order to better match,the recognition rate of the system is improved,and the vein is calibrated through the branch nodes of the vein.At the same time,the extracted SIFT feature vector is binarized and dimension-reduced,the feature vector size is greatly compressed,and a good recognition rate is obtained.