Research on the Palm Print Recognition Algorithm Research Based on Improved Geometric Invariance Principle
Xianping Wu · INTERNATIONAL JOURNAL ON Advances in Information Sciences and Service Sciences · 2012
The paper studies the problem of identity authentication system based on palm print recognition. It puts forwards the fusion algorithm combining the geometric rotational invariance characteristic with the straight line characteristic to solve the negative problems of deformation and interference in biometric identity authentication system based on non-contact pickup, which effectively improve the accuracy of palm recognition. First, it matches the two features—SIFT and LOFO extracted from the images of palm and gets the match value, and then uses the fusion algorithm to organically integrate these two matching results and gets the final match value. The advantage of this method is that it can more effectively afford the recognition in dealing with the negative factors such as rotation and scaling. And it uses the Gabor filtering algorithm to treat the original image adopted in the algorithm which can effectively extract the richer feature information and help a lot in the final recognition. The method can be used for personal identification and authentication.