Partial Differential Equations-based Fingerprint Image Enhancement Technology

Automated fingerprint identification system is the main research topic of biometric recognition. However, the performance of fingerprint identification systems is limited by the quality of fingerprint images which are often influenced by different factors. So it is necessary to improve the fingerprint image quality and to make them fit for the requirement of real applications. The preprocessing of the fingerprint images was discussed in this paper and much more attention was paid to fingerprint image enhancement with nonlinear partial differential equations.At first, we discussed the traditional methods of fingerprint preprocessing such as image normalization, the computation of orientation field, Gabor filters based image enhancement, fingerprint image binarization and image thining. Based on these works, we provided a good foundation to our further research work.Based on above work, a pseudo-linear directional diffusion filter is proposed to enhance fingerprint images according to the directional information of them. The new algorithm outperforms the coherent diffusion based image enhancement in structure information preservation and its pseudo-linear strategy decreases the computation burden of nonlinear partial differential equations to some extend.Because of the nature of the low-pass filter of pseudo-linear diffusion filter, the new method decreases the contrast between ridges and valleys in fingerprint images. In order to retain the contrast, we proposed to couple shock filters into pseudo-linear diffusion filter above. The coupled shock filters will make up for the deficiency of nonlinear directional diffusion based methods.In the last part of this dissertation, we summarized our work and analyzed the improvements to be done in the future.