Automated Fingerprint Identification System, Research And Implementation, Of The Minutiae Extraction Algorithm

The problem of resolving the identity of a person is critical in a variety of applications in our electronically interconnected society. Biometrics, as a reliable approach to positive identification, is based on identifying the person's physical or behavioral characteristics. Since the biographical characteristics cannot be easily forgotten, shared or misplaced, Biometrics is increasing deployed in authenticating the identification and detecting the identity crime or fraud. Among all the biometric techniques, fingerprint-based authentication systems have received the most attention because of the long history of fingerprints and their extensive use in forensics and civil applications, such as criminal identification, access control and ATM card verification. Though rather different from each other, almost all automatic fingerprint identification systems (AFIS) operate in two stages: minutiae extraction and minutiae matching. In this thesis, a novel fingerprint minutiae extraction algorithm is proposed to extract the minutiae from gray-scale fingerprint images directly. Firstly, a brief overview is presented to biometrics and biometric technologies. We mainly focus on the introduction of fingerprints techniques, including the history of fingerprints and its application in the authentication systems. Then, the fingerprint minutiae algorithm is described in detail. A new fingerprint segmentation method is used in the fingerprint pre-processing, while a new fingerprint classification method is applied to classify the fingerprints. Based on the directional information, the ridgelines in fingerprints can be followed to get a skeleton fingerprint image. After a post-processing procedure removing the false minutiae structures, it will be facile to detect the true minutiae from the skeleton fingerprint images. Experimental results can prove the efficiency and robustness of our algorithm.