The Research And Implementation Of Prototype System Of Automatic Fingerprint Identification

With the development of the science and technology, more and more high-tech electronically products come into people's daily life. In the same time, personal identification also becomes a problem which always appears in people's daily life. Then the biometrics technologies spring up gradually and become a kind of new personal identification method. Now, biometrics has been developed including fingerprint recognition, handprint recognition, Iris recognition, signature recognition, speech recognition etc.Fingerprint recognition has many advantages compared to other biometrics. It's practical and feasible. Minutiae-based matching is the main method of fingerprint recognition and most fingerprint systems currently in use are based on this kind of methods.Based on the above mentioned achievement, we designed and developed an Automatic Fingerprint Identification Prototype System. What we have done are as follows:(1)This paper introduces preprocessing methods of fingerprint image in detail. The process of image preprocessing includes normalization, acquirement of orientation image, filtering enhancement, binarization, thinning etc.(2) The Gabor filtering algorithm is applied in fingerprint enhancement.(3) We study and implement thinning algorithm based on morphology.(4) We implemented the methods of feature detecting and matching.(5) We create a small scale of fingerprint database. The performance of the proposed algorithms is evaluated through the database.Because our fingerprint database is small, our experimental results do not achieve the universal sense. However the experimental results indicate that automatic fingerprint recognition algorithm in the paper is effective and attains the requirement of practicality.Automatic Fingerprint Identification System integrates many disciplines, such as biologic characters, image processing, database etc. It has a great value in both practicality and theory.