Automated Fingerprint Identification System Algorithm

With the rapid progress of computer science and network technique, an accurate and reliable personal identification technology becomes more and more important. Biometrics based on personal identification methods have received extensive attention. Fingerprint recognition technique is the earliest one and is applied widely in the all biometrics recognition techniques. With the increasing requirement, people also want the automatic fingerprint recognition system become more reliable. The application of technology does not prevail at present, because the accuracy and speed of recognition is far from satisfactory to many practical circumstances. To improve the performance of automated fingerprint recognition is necessary.The dissertation focuses on automatic recognition of fingerprint. Based on current theoretic framework, a complete algorithm of fingerprint recognition has been accomplished on Visual C++. Attention have been given to some problems for further improvement. The main contributions of our work reported in this thesis are as follows:1. Based on research of recent segment algorithm, an algorithm combining variance and pixel clustering has been introduced. The algorithm which decide front block block according to proportion of front pixel and background pixel remedy the shortage of single threshold segmentation.2. improved Methtre algorithm has been introduced. The dot orientation of pixel has been get by calculating dot orientation, average and variance.3. A orientation match algorithm has been introduced which filter orientation block of fingerprint image. The algorithm which build filter window for eight orientation, filter orientation block by orientation match enhances the fingerprint image.4. Traditional post-thinning algorithm is improved by holding furcation and deleting bifurcate joint. The algorithm decrease false minutia and increase veracity of minutia extracting. Finally, summarize the work of the thesis, analyze the improvements need to be done and give the direction of future work.