Apply Bi-Features to Implement a Palm Vein Recognition System and Device Based on Gabor-Min
| Autor: | Yo-chen, Cheng, 鄭又誠 |
|---|---|
| Rok vydání: | 2013 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 101 One of the most enlightened subjects in science field these years is Biometric – which recognizes one’s identity through either his physiological or behavioral characteristics. And our research in adopting the information of palm vein is the realization of Biometric. One person’s vein information mostly remains stable in its pattern as growing older, and carries its unity. Also, because this information is ingrained deeply in a skin, the information for Vivo Identification, it could not be possibly duplicated thanks to its high occultation. Therefore, Palm Vein Feature is regarded as the most potential and advantageous option out of many Bio-Characteristics. In this study, we come out of a whole new design which is able to firmly fix one’s palm position and the degree of (palm) muscle stretch, providing a better detecting environment of identification. Also, in the palm vein identification system, we put forward the two-phased multiple feature identification – Harris corner vector of CCH (Contrast Context Histogram), and the SURF (Speeded Up Robust Features) vector based on Gabor-Min texture reconstruction. This system also creates a mechanism of updating historical records, which helps increasing the accuracy of identification and the stability under long-time use. Proved by experimental results, and to compare with former thesis, this thesis including hardware design and the proposition of identifying method in it, truly has advantages of higher recognizing ratios and lower costs. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|