Spoof Detection of Fingerprint Biometrics using PHOG Descriptor

Biometrics are used for authentication. It isused to recognize a person based on their unique characteristics. Among several biometrics, Fingerprint is the most widely used and acceptable biometrics. Biometric system has several advantages over traditional methods. But it can be affected by several attacks. In this paper type 1 attack is discussed which is performed at the sensor level. Differentiating a genuine biometric input from fake input is known as Spoof detection. The local features of fingerprint are extracted using Pyramid Histogram of Oriented Gradients image descriptor. It is used to detect whether fingerprint is real or fake.