IRIS Spoofing through Print Attack Using SVM Classification with Gabor and HOG Features

The main objective of this work is to create an easy-to-use and effective mechanism to help in preventing iris spoofing attacks by devising a system which tells us whether the scanned item is a real iris or a fake one. To overcome and prevent this situation, we have devised a system which helps us in preventing iris printout attacks by using feature extraction methods of Gabor filters and HOG bins (Histogram of Gradient). Post that for the purpose of classifying, we use Support Vector Machines (SVM) as a classifier in the designed mechanism. A proper implementation of this mechanism can help in detecting print attacks of iris at various security levels, and in an effective and user friendly way. It is not only capable of operating with a very good performance under different biometric systems (multi-biometric) and for diverse spoofing scenarios, but it also offers a high security against certain non-spoofing attacks (multi-attack).