الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 140 |  |  |
| | | from | 140 | | |
Abstract
There are many trails to increase the accuracy and speed of biometrics recognition systems. In this thesis, we presented a method to integrate more than one biometric to the same person for increasing the accuracy. 2D: 4D ratio computation has been used to increase the speed of biometric recognition. This ratio used to add each person to his class according to the ratio value, which led to a reduction in the number of comparisons between input and registered images. In order to preserve the used biometrics that represent real parts of person and prevent illegal using, this thesis presents an attempt to generate a piece of fingerprint instead of capturing and storing it. The proposed system for applying an identification system for identifying the person’s identity depending on using face biometric. The idea here is that no needs to touch or impact with any device. Face biometric was chosen. However, face biometric is very important biometric for identifying personality, but using unimodal biometric system becomes less efficient because of intra class and inter class issues. To increase system security we proposed a second phase to build a multimodal biometric recognition system called impact system. The biometric science depending on storing biometric templates after applying enrollment phase to create the system database. In reality theses stored templates represents a real parts of person, which can be used in bad ways. Even the biometric owners died their templates are still available in the stored templates from enrollment phase. So a lot or researchers start to protect these templates and create databases by encrypting it in the enrollment phase. As there are ways to protect these databases, there are many ways to hack it or stole it or use it badly. All of the aforementioned causes necessitate us to find methodology to generate fingerprint patterns instead of storing it’s template for using in matching phase of biometric recognition systems. Then we proposed a start trail to generate piece of fingerprint instead of capturing and storing it.