An Object-Face Liveness Detection based on Local Feature Descriptors Using Fuz-SVM Classifier

摘要

In day to day life, the Object Face liveness detection of specific image became a critical task in user authentication and other security applications. However, it is important to develop a system which can improve in processing of user authentication and also accomplished with identification of spoofing object from its genuine objects. In practically, most of the existing systems were fails in distinguish of genuine and fake objects especially in case of blur images. In order to overcome this situations and identify genuine object live faces, we proposed An Object Face Liveness Detection based on Local Feature Descriptors using Fuz-SVM Classifier. In the proposed system we used Histogram of Oriented Gradients (HOG) combined with Local Phase Quantization (LPQ) so that the HOG will extract features from normal images and LPQ will extract from blurred images. The proposed system also concentrates on individual differences among several objects, allows to select specific part of whole object based on Region of Interest (ROI). One of the major advantage with this system is reduces the processing time, and increases recognizes rate of specific face of the object.%Liveness Detection; Object-Face; Authentication; Original Objects; Fake Objects; Region of Interest; Local Feature Descriptor; HOG-LPQ; Fuz-SVM classifier