Palmprint Recognition Based On Local Joint Edge And Orientation Patterns

In the era of network information,the problems on how to identify one person and protect information effectively are the key social issues to be solved quickly.Biometrics uses the inherent physiological or behavioral characteristics of the human body to identify individuals.Because of its security,reliability,simplicity and convenience,biometrics has gradually replaced traditional authentication technologies based on possessions or passwords.Palmprint recognition is a relatively new biometric identification technology that has many unique advantages.Compared with the fingerprint,palmprint contains abundant texture information in a bigger feature regions.Compared with face recognition,palmprint recognition can distinguish twins more precisely.Compared with iris recognition,palmprint recognition can be accepted easily by users because the palmprint image is easy to capture.Compared with vein recognition,palmprint recognition has higher recognition rate and high reliability.However,palmprint recognition technology has not been well-developed yet.In order to design a practical automatic palmprint recognition system,more in-depth researches are needed.This paper introduces the basic framework of palmprint recognition system and proposes an effective palmprint recognition method based on local descriptors.The main work of the thesis is as follows:(1)Double Gabor Orientation Weber Local Descriptor for Palmprint Recognition.In order to improve the palmprint recognition rate,the differential excitation and gradient orientation of weber local descriptor(WLD)are improved based on the texture features of palmprint images,and double Gabor orientation weber local descriptor(DGWLD)is proposed.The directional information of the difference between the neighborhood pixels and the central pixel is considered to enlarge the difference between palmprint,when constructing the new differential excitation map.At the same time,gradient orientation is replaced by double Gabor orientation to reduce the influence of translation and rotation.In addition,a feature cross matching algorithm is used for further improve the recognition rate.Experiments on PolyU,PolyU MSpalmprint and CASIA palmprint databases show that the recognition rate is up to 100%.The experimental results show that the proposed method is superior in terms of identification rate and equal error rate compared with other local descriptor methods and improved WLD methods..(2)Multispectral Palmprint Fusion Recognition Based on Local Joint Edge and Orientation Patterns.In view of prominent edge and orientation feature in palmprint images,a new local joint edge and orientation patterns(LJEOP)is proposed to extract the palmprint features.Firstly,the Kirsch operator is used to calculate the edge response of the palmprint image in eight different directions,and maximum number of edges is used to describe the edge features of a palmprint image.The Gabor filter or the improved finite Radon transform(MFRAT)is used to extract the orientation feature.Secondly,the two-dimensional feature matrix is constructed based on a combination of edge and orientation analysis.Experimental results show that the LJEOP method has higher recognition rate,lower equal error rate and faster recognition speed than state-of-the-art methods.Finally,the matching scores of each spectrum are weighted by equal error rate for the higher recognition accuracy.