|
[9002986.] رقم البحث : 9002986 - |
|
Representation Learning Framework of Object Recognition via Feature Construction / |
| تخصص البحث : |
| |
Mansoura journal for computer and information sciences / / Vol.13 - No.1 |
| |
Shaimaa. A.M. Hegazy ( shaimaa_hegazi88@yahoo.com - ) - مؤلف رئيسي |
| |
Mostafa G.M.Mostafa ( mgmostafa@acis.asu.edu.eg - ) |
| |
Ahmed Abu Elfetouh ( elfetouh@gmail.com - ) |
| |
Biometric system, SDUMLA Database, Iris Recognition, Daugman’s Rubber sheet, Haar Wavelet Transformation (HWT), Principal Component Analysis (PCA), Euclidean Distance (ED). |
| |
Biometric technologies are very important these days for improving the accuracy of protecting private data from unauthorized access. It helps overcome deficiencies of current security traditional systems. For the last decade, researchers are developing new methodologies that employ biometrics to boost security field. This article proposes effective methods for Iris recognition based on multi-feature fusion.
A feature fusion approach is implemented to improve the iris recognition rate. In particular, Haar Wavelet Transformation (HWT) features and principal Component Analysis (PCA) are used to model the iris texture. Both approaches are fused to improve performance. Fusion results are compared to those from each feature alone and with other reported work. The results obtained with the proposed method are better than the currently reported results. |
تسجيل دخول 
