Research On Non-contact High Quality Palm Vein Image Acquisition Method

The rapid development and extensive application of information technology has brought great convenience to people's lives.While enjoying a convenient and comfortable life,we are also facing increasingly serious information security threats.Therefore,secure and accurate identity authentication technologies have become Research hotspots.Palm vein recognition technology has the characteristics of living recognition,internal features,non-contact,high security level,and has received more and more attention in recent years.This article mainly focuses on high-quality palm vein image acquisition methods.The research includes the design of palm vein acquisition equipment,rapid palm positioning and region of interest extraction,rapid automatic palm vein image quality assessment,and palm vein image enhancement.The specific research results of this paper are as follows:A variety of methods for locating the palm were compared and a new rapid palm positioning method was proposed.In this paper,the palm positioning method based on the distance between the center of the palm or the wrist and the contour of the palm is introduced,and some improvements are made to the method,so that the algorithm can extract the region of interest without judgment of the right and left hand.In addition,a palm positioning method based on palm profile concave analysis is proposed,and it is proved through experiments that the method has a very high positioning accuracy.After acquiring the palm's region of interest,an image quality evaluation method based on the gray mean difference between patches is proposed to evaluate the illumination uniformity of the region of interest in palm vein image.The image quality evaluation method based on Reblur was improved to make it more suitable for palm vein images.Inspired by the natural scene statistical image quality assessment algorithm,a palm vein image quality assessment method based on the statistics of Haar-Like feature was proposed.The effectiveness of each algorithm was verified by experiments,and fusion of multiple algorithms was selected to evaluate the quality of palm vein images.Finally,a new method of contrast enhancement is proposed.Through contrast experiments,it is determined that the Gaussian fitting gray transformation algorithm proposed in this paper can enhance the contrast of palm vein ROI images.Finally,using the palm vein acquisition equipment designed in this paper,combined with the palm positioning algorithm,palm vein quality assessment algorithm,contrast enhancement algorithm and the existing feature extraction and matching algorithm,the entire non-contact palm vein recognition system is completed.The accuracy of the system's identification was tested through experiments.The system's FRR was as low as 2.5%when the FAR was 0.01%,which meets the requirements of practical applications.