Development of a Hybrid Platform for Thumbprint Forensics

Abstract

Thumbprint has remained the most commonly used technique for biometric identification in the world today. Research activities on the improvement of the theory and practice of thumbprint forensics have, therefore continued to be a major task in the world today. The research which is reported in this paper takes a study of the features of the existing and popular research activities on the theory and practice of thumbprint forensics, analyzed their Strengths, Weaknesses, Opportunities and Threats (SWOT) and proposed a hybrid system for thumbprint forensics which integrates the strengths and opportunities. The hybrid system is characterized by an architecture which is composed of database of thumbprint profile, data mining engine and decision support engine. The data mining engine is composed of thumbprint image enhancement, thumbprint minutiae extraction, database of thumbprint minutiae extraction, thumbprint pattern recognition and thumbprint pattern matching. The decision support engine is composed of error detection and error correction subsystems.

Keywords: Forensics, Biometrics, Thumbprint, Minutiae, Segmentation, Normalization, Binarization