Font Size: a A A

Research On Palmprint Feature Extraction And Pattern Matching Algorithm

Posted on:2017-04-24Degree:MasterType:Thesis
Country:ChinaCandidate:Z L MeiFull Text:PDF
GTID:2348330488996110Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Biometric identification technology has won the widespread attention of people because of the characteristics of its safety and efficient,capacity and accuracy,broad application.Palmprint recognition is a new biometric identification technology developed in recent years,palmprint recognition not only has the common feature of biometric identification technology,but also has some unique advantages,such as rich and stable feature information,easy for users to accept,less influenced by external conditions,and low price of acquisition equipment.Therefore,palmprint recognition has become a research hotspot in biometric identification,and the research of palmprint recognition has great significance in the field of information security,public security and so on.The core issue of palmprint recognition is the recognition speed and accuracy,this paper studies the two core issues of palmprint recognition,the main work and contributions are as follows:(1)Introducing biometric identification technology compactly,summarizing the significance of the biometric identification technology,expounding the system of biometric identification technique,and parsing the development prospect of biometric identification technique,introduces the several current representative biometric recognition technology emphatically.Palmprint recognition technology is introduced from the aspects of palmprint pretreatment,feature detection and matching identification.(2)A palmprint identification technique on the strength of recusive Gabor filter is proposed in the paper.Traditional Gabor,improved Gabor and recusive Gabor are used in palmprint recognition feature extraction,in palmprint matching recognition stage competition coding method is used,calculate the angle and distance to accomplish the recognition,study their problems of feature extraction speed in palmprint recognition comparatively,to seek a faster palmprint recognition method with good recognition accuracy.(3)A method for palmprint recognition based on support vector machine SVM(support vector machine,SVM)and Hamming distance HD(Hamming distance HD)is proposed in terms of palmprint recognition accuracy.Using morphological operators to locate and segment the palmprint image ROI(region of interest,ROI),decomposing the ROI with Haar wavelet,then extract and code the low frequency sub image which comes from the decomposition of the ROI with 1D Log-Gabor filter,study the accuracy of several matching modes comparatively,to seek a higher accuracy palmprint recognition method with good recognition speed.
Keywords/Search Tags:Biometrics, Palmprint recognition, Gabor filter, Support Vector Machine, Hamming Distance
PDF Full Text Request
Related items