Research On Identification Technology Based On Finger Vein Image

Finger vein based identification technology has become a hot topic in many biometric recognition technology fields.This article mainly studies the methods of finger vein image collection and recognition,mainly including:introduction to the background and significance of finger vein feature recognition technology;Introduction and design of vein imaging system;research and optimization of finger vein image processing algorithms;experimental verification and simulation implementation of identity recognition system,as follows:1.Introduction to the research background and significance of finger vein recognition technology.Based on a comprehensive introduction to several common biometric technologies,the current status of finger vein recognition technology at home and abroad and the advantages of using it for identity authentication are highlighted.2.Design of finger vein imaging system.Based on the research on several typical finger vein collection systems,a double light wave vein collection system is designed in this paper.The test shows that the dual light wave vein acquisition system designed in this paper can handle the acquisition requirements under different conditions on its own,and ensure that the captured finger vein image has uniform brightness and clear texture.3.Research and optimization of finger vein image processing algorithms.Based on the analysis of several typical finger vein image recognition algorithms,this paper proposes an improved guided image filtering algorithm,a new OPTA thinning algorithm,a unique dual competition mode fusion method,and optimization of finger vein image processing such as sparse representation.algorithm.4.Experimental verification of the identification system.In the recognition system authentication phase,we combine the algorithms proposed in this article with PCA,PCA + LDA,PCA + SRC classic algorithms in the finger vein image database of Shandong University,the finger vein image database of Hong Kong Polytechnic University and the dual light wave vein collection designed in this paper.Experimental verification is performed on the database collected by the system.The experimental results show that the recognition algorithm proposed in this paper has higher recognition rate and better stability on three sets of data sets than the other three classic algorithms,and the recognition rates are all above 90%.