Comparative Analysis of Fingerprint Orientation Field Algorithms

Fingerprint images are textural images consisting of ridges and valleys. The orientation of textures can be determined by orientation field computation. Fingerprint orientation field is the critical basis for fingerprint image segmentation, filtering enhancement and matching processes, and the fingerprint orientation field algorithm plays a very important role in the applied Automated Fingerprint Identification Systems (AFIS). The available fingerprint orientation field algorithms include mainly the mask algorithm and the gradient algorithm. They are used in spatial and frequency domains, respectively, and both can offer satisfactory fingerprint orientation matrixes. Based on an experimental comparison of their effectiveness in fingerprint preprocessing, this paper analyzes in detail the performance of these two algorithms.