Research And Realization Of Fingerprint Identification System

Fingerprint recognition is one of the earliest technologies for using biometrics for identification,and it is also a very mature technology.The fingerprint recognition algorithm is mainly divided into two parts: fingerprint image preprocessing and image feature point matching.Since the extraction and matching of fingerprint feature points is based on the preprocessed image,fingerprint preprocessing has a significant impact on the accuracy of fingerprint image recognition.It can even be said that the accuracy and reliability of fingerprint feature extraction directly depend on the quality of fingerprint preprocessing,and fingerprint preprocessing directly determines the final effect of fingerprint recognition.To propose new methods and techniques to solve the shortcomings in current fingerprint preprocessing,the preprocessing technology of fingerprint images has been deeply researched in this paper.In the relevant chapters of this paper,image preprocessing algorithms including fingerprint image normalization,grayscale equalization and fingerprint image segmentation are studied.In this paper,an improved Gabor filtering algorithm is used,which uses the direction and frequency information of the fingerprint to achieve better image enhancement.Based on the analysis of fingerprint image binarization,a binarization method based on gradient direction information and frequency information is selected to reduce the image quality requirements of gradient.In the research of the basic fingerprint image refinement algorithm,an improved OPTA refinement algorithm is selected,which can effectively suppress the glitch and improve the refinement effect.After the above research results,a relatively mature identity authentication system has been completed based on fingerprint recognition.The fingerprint is used as a way for the user to log in,which overcomes the defect that the traditional user logs in using the personal password.We first carried out functional and non-functional requirements analysis,and according to the analysis above,designed the main functional modules of the fingerprint identification system and engineered the practitioners,and then combined the modules through the MFC framework to construct a visual fingerprint identification system.Finally,the fingerprint identification system is tested and applied.However,due to conditions,the performance of large-scale system applications could not be verified.