Reference Point Determination in Enhanced Fingerprint Image

Accurate reference point detection is crucial for features extraction to improve the performance of fingerprint recognition system. Many reference-point detection algorithms work well in good-quality fingerprints, however, most fail to correctly locate the reference point in poor-quality fingerprints with cracks and scars, dry skin, or poor ridge-and-valley contrast. In this paper, we propose a method to detect a reference point in an enhanced fingerprint image by using a novel and effective enhancement approach. It preserves true ridges and valleys, and it also repairs the damaged ridges and valleys, which in turn helps to detect reference points reliably and accurately. Experimental results on public databases show that the proposed method has better results when a reference point detection method applies to the enhanced image, than to the original image.