Fingerprint image enhancement on edge detection techniques using Scilab

Abstract

Nowadays, fingerprint becomes an important thing in security system, replacing the passwords, however the unclear image of the fingerprint itself brings a negative effect to the user, there could be various problems related to fingerprint image discontinuities such as skin problems, the texture of the ridge and lighting conditions. To improve the quality of fingerprint image, it needs to undergo an enhancement process based on the edge detection technique. Hence, the purpose of this project is to investigate and identify the suitable algorithm for edge detection techniques of fingerprint images. The Histogram equalization, Binarization and Edge detection technique integrated with the Robert, Prewitt, Sobel and Laplacian algorithm are used in this project. The performances of each edge detection technique are compared based on the intensity versus edge sharpness to determine the most preferable edge detection techniques to solve the problem of edge discontinuities using Scilab. From the simulation, it found that the Laplacian operator is the best technique to detect the edge of fingerprint images.