The research on image enhancement of low-quality fingerprint image

Abstract

Fingerprint Identification is one of the most reliable and popular personal identification methods. It can be divided into image acquisition, enhancement, feature extraction and matching steps. The enhancement stage is of great importance as it influences the performance of subsequent feature extraction and matching. In this paper, different image enhancement approaches presented in the scientific literature are reviewed. Many fingerprint enhancement algorithms were effective for medium or high quality fingerprint images. But these existing algorithms were either not suited for low quality fingerprint images, or too slow to satisfy the real-time requirements. We proposed a fingerprint enhancement algorithm based on Fourier filtering. In our algorithm the fingerprint enhancement were transformed from spatial domain to frequency domain by Fourier transforming. The fingerprints were enhanced by band-pass filter and directional filter in frequency domain. Experimental results show our algorithm is fast and has excellent enhancing performance for low quality fingerprints. In addition, Enhancement for non-minutiae based fingerprint recognition for low quality images is presented. the pre-processing stage operating on a hexagonal grid can be implemented efficiently and gives promising results. Methods of comparing the performance of enhancement methods are discussed. Conclusions are made regarding the importance of effective enhancement, especially for noisy or low quality images.