Fingerprint databases consisting of 185 real, 90 fun-doh and 150 gummy fingerprints are created. Features are also tested on a hybrid classifier formed by fusing all the mentioned earlier classifiers by the |majority voting rule|. LBP features and wavelet energy features are independently tested on various classifiers: AdaBoost.M1, support vector machine and k-nearest neighbour. Dimensionalities of the feature sets are reduced by running sequential forward floating selection (SFFS). Wavelet energy features characterising ridge frequency and orientation information are also used for improving the efficiency of the proposed method. Local binary pattern (LBP) histograms are used to capture these textural details. It is based on the observation that, real and spoof fingerprints exhibit different textural characteristics. ![]() To alleviate these problems, in this paper, a new texture-based method which needs only one fingerprint is proposed. ![]() Some other methods in the literature need extra hardware to detect liveness. It requires two consecutive fingerprints to notice perspiration and therefore the method is slow. Article: Local binary pattern and wavelet-based spoof fingerprint detection Journal: International Journal of Biometrics (IJBM) 2008 Vol.1 No.2 pp.141 - 159 Abstract: Perspiration phenomenon is very significant to detect liveness of a finger.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |