To search, Click
below search items.
|
|

All
Published Papers Search Service
|
Title
|
Iris Recognition Based on Ripplet Transform Feature Extraction
|
Author
|
Daruosh Kavosi, Abas Karimi
|
Citation |
Vol. 17 No. 7 pp. 231-234
|
Abstract
|
A novel technique for iris recognition is proposed based on Ripplet transform. This method uses the feature given from Ripplet components. Firstly, the geometrical region of iris image is detected by segmentation using Hough transform, and then the normalization is done to produce the iris section. A set of main intrinsic directional of iris section are chosen to minimize the processing time of recognition for two-dimensional Ripplet transform. The main diagonal Ripplet components are selected. The extracted components converts to bits similar to demodulate the binary phase shift keying (BPSK) symbols. Finally, the produced iris code is generated and can be used for identification. Simulation results investigate that the proposed Ripplet technique increases the identification accuracy in comparison to fast Fourier transform (FFT) and discrete cosine transform (DCT) methods.
|
Keywords
|
Transform feature, ripplet, iris image
|
URL
|
http://paper.ijcsns.org/07_book/201707/20170733.pdf
|
|