An Adaptive Algorithm for Smoothing Fingerprint Orientation Fields

Abstract

The traditional algorithm for smoothing fingerprint orientation fields was based on the low-pass filter. To improve this algorithm, a variable window adaptive smoothing algorithm for the fingerprint orientation fields is proposed based on the fingerprint orientation consistency. Our algorithm uses the correctness of the fingerprint orientation field to compute the weighted average of the orientation field. Experiments show that our algorithm can effectively deal with the orientation field images which have noises and really describe the topological structure of the fingerprint ridge.