PATTERN ANALYSIS MODEL FOR THE INVESTIGATION OF FINGERPRINT INDIVIDUALITY

Abstract

The need for more robust, reliable and efficient systems for the investigation of fingerprint individuality has been on high demand recently. Areas of such need include Police, Security, Biometric, investigation of criminal cases, and so on. Although, various techniques have been developed to address the problems of fingerprint individuality using different models with specific Strengths and Weakness. This research is motivated by the need to address problems of high error rate, intra class variation, inter class similarity among others that are associated with fingerprint individuality. A singular point and pattern type platform for individuality of fingerprints is proposed in this research. The platform combines the strength and also reduces the weaknesses of some of the existing systems. The first subsystem of the platform is responsible for the acquisition of fingerprint while the second subsystem is responsible for the enhancement of fingerprint images. Extraction of the singular point and pattern type features takes place at the third subsystem and individuality investigation based on component matching and detection and correction of associated errors take place at the fourth subsystem. The implementation was carried out in an environment characterized by Microsoft Windows while Matrix Laboratory (MATLAB) and Microsoft Access served as frontend and backend respectively. A case study of benchmarked fingerprint and locally enrolled fingerprint databases was used to establish the practicability, adequacy and acceptability of the proposed system. With the threshold value of 95%, the study produced FNMR value of 3.50% and FMR value of 0.16% which revealed satisfactory performance of the model.