Fingerprint Image Quality Assessment And Preprocessing

Fingerprint identification is one of the most important and the most widely used biometric technologies. More attention is paid to key techniques of fingerprint identification such as fingerprint image enhancement, minutiae extraction, fingerprint classification and matching but research on the quality of fingerprint image is fewer. Fingerprint image quality assessment and preprocessing is discussed in this paper.In practical application, the performance of the AFIS(Automatic Fingerprint Identification System)relies heavily on the quality of the input fingerprint images. The poor images will lower the performance of the AFIS. So fingerprint image assessment is becoming a crucial part and has drawn a substantial amount of attention recently. In order to make the AFIS operate reliably and efficiently, it is deadly needed to control the quality of the input images. A fingerprint image quality assessment module is added to the system. We achieve this module both in space domain and in frequency domain. The results show that both of algorithms reflect the quality of the input images objectively, apparently, accurately.Fingerprint image quality assessment paves the way for the following processing such as fingerprint segmentation, fingerprint enhancement, minutiae extraction. Based on the character of the spectrum of the fingerprint image, we put forward a new method based on FFT(fast Fourier transform)to segment the fingerprint images. Canny algorithm is used to detect the edge between the background and foreground of the fingerprint images, based on this, we put forward another segmentation algorithm. Based on description for an image used DT(Delaunay Triangle), a novel method of fingerprint segmentation is proposed. After calculating the orientation and the frequency of the fingerprint image accurately, Gabor filters which have both frequency-selective and orientation-selective properties are applied to enhance the fingerprint images.