A high precision recognition method for small area fingerprints based on machine vision

Abstract

Aiming at the problem that the traditional small-area fingerprint recognition method is insufficient to recognise the feature points in the boundary region and the recognition accuracy is low, a high-precision small-area fingerprint recognition method based on machine vision is proposed. Firstly, by analysing the estimated values of key fingerprint parameters, Tico descriptor is introduced to obtain detailed feature points and determine the frequency field. Then, the fingerprint image of small area is enhanced based on direction and frequency to make the image features clearer. Then, based on the enhanced fingerprint image, the fingerprint model is demodulated by a suitable two-dimensional signal, the detailed features of the small-area fingerprint image are extracted, and the direction vector Angle is doubled to achieve direction smoothing, so as to achieve a better feature representation. Finally, high-precision identification of small area fingerprints is realised by matching the fingerprint point set. The experimental results show that the method proposed in this paper can extract the detailed features of small-area fingerprint images more accurately, the recognition results are more accurate, and the average recognition time is 32.6 s, which can improve the recognition efficiency, and has certain advantages.