Edge detection and feature extraction in automated fingerprint identification systems

Abstract

As a means of access control and criminal identification, Automated Fingerprint Identification Systems (AFIS) are widely utilized. In many areas of business, these systems are entrusted to verify identities of personnel before allowing access to restricted information or facilities. In the area of criminal investigation, these same systems are entrusted to find, match, and identify criminals. Obviously, these systems are given critical tasks and are performing them unsupervised most of the time. Although most steps in the process are procedural and can be automated, there are two critical phases that need to be performed intelligently and reliably. These two phases are edge detection and feature extraction. In order to enhance important features accurately in the fingerprint image, methods in edge detection are applied. Once the important features are exposed and artifacts are removed, feature extraction takes place. In this phase, the print is characterized by the extracted features for matching later. This article looks at the theoretical foundations and practical aspects of these two phases to understand their automation.