An automatic fingerprint identification system (AFIS) has become one of the most widespread biometric identification systems for security. In this thesis, we introduce a minutiae-based automatic fingerprint identification system by using a graph point pattern matching algorithm, weighted maximum bipartite matching, to compute the matching score by comparing the minutiae between two fingerprint images. Our system is tested on 4 databases of fingerprint images, such as Rindex28, Lindex101 from PRIP Lab at NTHU, FVC2000 DB1 and FVC2002 DB1, on a Pentium 4 with 3.4GHz CPU and 2.5GB SDRAM running Windows XP. According to the type of top 3 matches, the recognition rates are 100% (112/112), 92.3% (373/404), 87.5% (70/80) and 52.5% (42/80), respectively.