Call For Paper Volume:7 Issue:9 Sep'2020 |

Secured Authentication with Sclera Veins by Wavelet and Feed Forward Neural Networks

Publication Date : 29/10/2016


DOI : 10.21884/IJMTER.2016.3099.7HGJX

Author(s) :

Anuradha R. Kondelwar.


Volume/Issue :
Volume 3
,
Issue 10
(10 - 2016)



Abstract :

Personal authentication nowadays had become the most crucial security reason. Various human features and signals have been used for identification along with hybrid approaches where more than one features and signals are combined for robust security reasons. This paper proposes a novel secure technique of human identification using the veins of sclera portion of a human eye. A 32x32 region is extracted for identification over the sclera part. A unique code is also generated for the 32x32 array for dual identification. Minimum 3 images are taken for an individual into database for training feed forward neural network with Backpropagation and 2 images for testing. The database considered for this implementation belongs to UBIRIS dataset. 12 different individuals are considered to test the validity of the proposed system. Debauchees 6 mother wavelet is used for final feature extraction over the 32x32 grid by converting it into a row vector. The final feature vector, after 4 level wavelet decomposition was a 30x1 array for each individual. The matching rate was 100%.


No. of Downloads :

2


Indexing

Web Design MymensinghPremium WordPress ThemesWeb Development

Secured Authentication with Sclera Veins by Wavelet and Feed Forward Neural Networks

October 15, 2016