Low-quality Fingerprint Image Enhancement And Deformation Fingerprint Matching

Fingerprint recognition is a technology of biometric authentication as well as the most ideal one. To improve the performance of fingerprint recognition and accelerate its prevalence, This thesis has studied the main technologies of fingerprint recognition, mainly including enhancement of low quality images, recognition of small fingerprints and match of distorted fingerprint images, and makes the following contributions:(1) Enhancement of low quality fingerprint images is analyzed, and a Log Gabor filtering based approach to fingerprint image enhancement is proposed. Gabor filtering is by far the most popular method for fingerprint enhancement, whereas there are still certain limitations. We propose a novel approach to fingerprint image enhancement, which is based on Log Gabor filtering. Ridge frequency and ridge orientation are used to establish Log Gabor filter. To implement the filtering in frequency domain, the windowed Fourier transform is utilized to extract the local frequency spectrum of fingerprint images. Experimental results show that the proposed algorithm effectively improves the quality of fingerprint image and promotes the veracity of fingerprint recognition.(2) For small fingerprint, a texture matching method base Log Gabor filtering is proposed. Gabor based texture matching was proposed to overcome the difficulty of small fingerprint recognition and achieved comparatively favorable performance. However, compared with Gabor filters, Log Gabor filters is more capable of texture analysis. A texture matching method based on Log-Gabor filtering is proposed. A quick and effective method of reference point location is utilized. According to center point feature region is extracted from fingerprint image and normalizing is implemented. Before Log Gabor filtering fingerprint images of feature region are transformed into frequency domain. Finally FingerCode is extracted from filtered image for matching. Experiments show that this approach performs better than Gabor based method.(3) A minutiae relationship representation and matching method based on curve coordinate system is proposed. For each minutia, a curve coordinate system is established, and the coordinates of other minutiae in this coordinate system is computed. Thus, the coordinate relationship between each pair of minutiae can be evaluated. These relationships are used for pairing minutiae between the template fingerprint and the query fingerprint by means of transferring reference minutiae. The results show that the proposed algorithm achieves improved matching accuracy and is able to cope with highly distorted fingerprints.