Research On Image Preprocessing Of Library Fingerprint Management System

With the continuous development of science and technology，fingerprint recognitiontechnology has been recognized as personal biometric identification physical evidence of thefirst. The Library Fingerprint Management System which is designed by the library of HebeiUniversity of Architecture uses comprehensively in September2006.According to the uniqueand stability of person’s fingerprint, the system combines computer technology, biologicalrecognition technology and artificial intelligence technology organically. By embedding in thelibrary management system of automatic fingerprint identification system, the system realizesthe readers’ undocumented management. Application of the library of fingerprint managementsystem has enabled the readers to borrow books conveniently and the library staff to reducethe workload and improve work efficiency. But, in the practical application process, theimages of the reader fingerprints are commonly of poor quality，which makes imageprocessing rather difficult and performs as a bottleneck to further develop the libraryfingerprint management system.This thesis briefly introduces the Library Fingerprint Management System，and then thefingerprint image preprocessing is studied.（1）The fingerprint image segmentation algorithms have been studied. This thesisintroduces the segmentation algorithm of fingerprint image based on gray variance and putsforward the improved segmentation algorithm. Improved fingerprint images segmentationalgorithm utilizes the texture features of fingerprint image, the segmentation effect of lowquality fingerprint image is good.（2）The fingerprint enhancement algorithms have been studied.This thesis study on thetraditional fingerprint image enhancement algorithm based on Gabor filter and the algorithmis improved. Improved Gabor filter algorithm combines with local and macroscopic area offingerprint image.It not only uses of the local characteristics of fingerprint image,but alsocombines with the characteristics of regional local neighborhood association of the fingerprintorientation upheaval area. Improved Gabor filter utilizes the ridge fixed frequency, it reduces time of calculating fingerprint frequency and improves the efficiency.（3）The fingerprint image thinning algorithms have been studied. This thesis introducesthe improved OPTA and puts forward improved fingerprint image thinning algorithm.Improved fngerprint thinning algorithm increases the elimination template and the modifyingtemplate.Improved thinning algorithms can achieve the ideal effect in the processing of thefingerprint image.Finally，verifying the improved algorithm though the Library Fingerprint ManagementSystem. The experimental results show that recognition rate of the Library FingerprintManagement System has markedly improved.