This paper presents a suggested prototype for developing an integrated system for person identification using fingerprints. This system encompasses three stages: The first stage dealing with the purpose of identifying the center point and developing a normalized image for the region surrounding the detected center point. In this respect, we have suggested a novel algorithm that was tested, compared with other established systems, and proven to be faster and more accurate. The second stage is the feature extraction of the fingerprint where a set of Gabor filters were used to extract the local and global features and transform them to a relatively short code with fixed length and uses it in the process of fingerprint identification. The final stage includes the identification of the fingerprint using a neural network that was trained using the back-propagation algorithm. The proposed system was evaluated over a live-scan fingerprint database. The proposed system was able to correctly identify an input fingerprint with an accuracy of (96.67%). On the other hand, the system did not mis-identify any input fingerprint.