Wednesday, September 30, 2020 Home   | Search
Viewing Abstract

Page: 433-442 ------------------------------> Last_modified :7/11/2017 12:11:00 PM

Title: Spoofed Iris Recognition: Synthesis of Gabor and LBP Descriptor Using SPPC
Authors: M. Malathy ; J. Arputha Vijaya Selvi
DOI:
Aff: Department of Information Technology, Kings College of Engineering, Pudukkottai, Tamil Nadu, India
Author Email:
Keywords: Biometric; Iris Recognition; Spoof Iris Image; Multi-Scale Local Binary Pattern (MLBP); Gabor Wavelet; Local Binary Pattern (LBP); Spatial Point Pattern Classifier
URLs: ABSTRACT-HTML  | FULLTEXT-PDF  | 
Abstract:Background: With the growing requirement of safety in everyday activities, identity management has quickly become an eminent concern. Biometrics using iris recognition systems are employed to scrutinizing an individual from a digital image or video source. In the proposed work, the robustness of iris recognition with spoofing attack is expounded. The dataset images are improved using image enhancement techniques. Multi-scale Local Binary Pattern (MLBP) algorithm is established for extracting the effective features to symbolize the images. This algorithm is a combination of Gabor wavelet followed by Local Binary Pattern (LBP) descriptor. The magnitude coefficients (real part) are isolated from Gabor wavelets and considered as an input for LBP which extracts the features of the given input image in various directional scales. Here, a supervised learning classifier especially spatial point pattern classifier is employed to check the verification process. Both dataset iris images and synthetically spoofed iris images are evaluated by the algorithm in order to increase a genuine acceptance ratio (GAR). The experimental result exhibits high efficiency and evaluates the better performance of the proposed approach. Objective: Spoofed Iris Recognition: Synthesis of Gabor and LBP descriptor using SPPC. Results: This section presents the result analysis for proposed work in which the spoofed iris image is detected by using the normalization method and feature extraction method. The classification accuracy and timing analysis of the algorithm are given in table I. In the presence of uncertainty, true positive rate or sensitivity is used to test the robustness of the results of a proposed system. Conclusion: In this paper, the robustness of iris recognition system with spoofing attack is explained. The dataset images are enhanced using image enhancement techniques. It is later subjected to multi-scale local binary pattern (MLBP) algorithm for extracting the efficient features to represent the images. This algorithm is a combination of Gabor wavelet followed by local binary pattern description (LBP) where the magnitude coefficient from Gabor wavelets takes as its input. A spatial point pattern classifier is used to check the verification process. Both dataset iris images and synthetically spoofed iris images are evaluated by the algorithm in order to increase a genuine acceptance ratio (GAR).,