Loading...
Please wait, while we are loading the content...
Similar Documents
Authentication System Depend on Hand Geometry Using Back Propagation Neural Network
| Content Provider | Semantic Scholar |
|---|---|
| Author | Al-Fiky, Firas M. M. Ageed, Zainab Salih |
| Copyright Year | 2014 |
| Abstract | In this method, the hand geometry measurement be used and a new features has been extracted from two landmarks determined at the first step of preprocessing . The images taken by scanner covered by a black box without peg. The data collected and then the pre-processing applied on it, after that passed to feature extraction stage. Each person has its own features like finger length, finger width, palm width, angle value and many other measurements. Features extraction contain 20 distances for fingers and palm (D1-D20) and 10 angels values measured and determine the center point on it . The angels calculated for a ten triangles determined, then calculate the Area and the Perimeter on it. The second part of the proposed method begin with training a neural network. The BPNN used here with trainlm algorithm after that testing the NN. The population size was 100 samples, for 10 person, each person have 10 samples. For Training set 70 samples taken and 30 for testing the NN. The Recognition Rate (RR) was 91%. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://www.jitbm.com/JITBM%2028th%20Volume/6zainab.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
