Anti-spoofing for display and print attacks on palmprint verification systems

摘要: A number of approaches for personal authentication using palmprint features have been proposed in the literature, majority which focus on improving matching performance. However, late, preventing potential attacks biometric systems has become a major concern as more and get deployed wide range applications. Among various types attacks, sensor level attack, commonly known spoof emerged most common attack due to simplicity its execution. In this paper, we present an approach detection display print based verifcation systems. The is analysis acquired hand images estimating surface re ectance. First higher order statistical computed from distributions pixel intensities sub-band wavelet coeefficients form feature set. trained binary classifier utilizes discriminating information determine if image real or fake one. Experiments are performed publicly available dataset, containing 1300 corresponding 230 subjects. Experimental results show that biometrics samples can be substituted by digital copies with alarming acceptance rate high 79.8%. also very effective between images. consistently achieves over 99% average 10-fold cross validation classification accuracy our experiments.