2D touchless fingerprint verification system

Abstract

In this project, a touch-less fingerprint verification system by using a web camera will be constructed. The system auto captures and processes captured fingerprint image extracting characteristic features to be compared with those of fingerprint templates for verification. Specifically, we address the constraints of the system, such as the low contrast between the ridges and the valleys in fingerprint images, defocus and motion blurriness. In order to compromise those constraints, the system comprises of preprocessing, image processing, feature extraction and matching stages. The web camera is calibrated to ensure a focused fingerprint image is captured. Several Region of Interest (ROI) sub-images of the captured image are selected to be segmented, filtered and enhanced for better image effect. Based on those sub-images’ variance and mean, best focused imaged is selected and captured for later feature extraction and matching use.