Font Size: a A A

Research On Palm Vein Image Acquisition Based On Optimized ROI And Feature Extraction Algorithm

Posted on:2019-04-27Degree:MasterType:Thesis
Country:ChinaCandidate:G LiuFull Text:PDF
GTID:2438330545490746Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
In the time of network information,people's requirements for information security are constantly improving.Identity authentication is the first barrier to protect data,but the traditional authentication methods,such as passwords,documents,etc.,are easily to be embezzled.In this context,biometrics came into being.Palm vein recognition technology is a kind of biometrics.This project designed a model of palm vein recognition,improved the problem that region of interest extraction and feature extraction algorithms have poor performance.Pretreat the original palm image,the preprocessing including the image background separation and smoothing the image.Aiming at the problem that the algorithm of palm vein feature extraction based on the maximum inscribed circle algorithm has too many traversal and inefficient execution,put forward an improved algorithm for extracting the largest inscribed circle.Add grid lines to the palm vein image as reference,reduce the size of the center,accelerate the process of determining the center.Determine the initial radius,change the radius of the inscribed circle with grid as the variable,determine the maximum inscribed circle.Aiming the local binary pattern(LBP)algorithm which the process of taking texture feature extraction exist a local,high dimension of the problem,put forward a feature extraction algorithm of palm vein image texture based on Gabor wavelet and improved LBP algorithm.According to the characteristics of palm vein,design a 5-scale 4-direction Gabor filter,access multiple amplitude characteristics.Divide LBP algorithm as blocks and reduced the algorithm in uniform mode to improve the LBP algorithm.Take the LBP algorithm to extract the texture features on each amplitude feature of the image.In order to prove that the improved algorithm can extract the palm vein characteristics with a better recognition performance,design experiment for feature matching recognition.Use three kinds of typical feature matching algorithms to match the palette texture features extracted by the proposed algorithm and the reference algorithm,then compare the statistical data.Experiment's data shows that the algorithm proposed in this paper can effectively improve the recognition performance of palm vein recognition.The algorithm of this paper has certain practical value.
Keywords/Search Tags:Identification, Palm vein recognition, Region of interest, Feature extraction, Feature matching
PDF Full Text Request
Related items