Research Of Fingerprint Identification Algorithm Based On Embedded Systems

With the rapid development of society, the needs of identification of individuals growing and everywhere. The performance of traditional identification methods limitations and drawbacks become increasingly apparent. The biometric identification technology which emerges quietly can essentially solve the traditional identification of technical deficiencies. It has gradually being recognized. Fingerprint identification has become the most reliable and user-friendly type of biometric identification technology. At present, the traditional PC-based fingerprint identification system has many shortcomings, for example, large, not easy to carry, poor real-time and so on. Thus the fingerprint identification system which based on the embedded has become a hot research now. In this context, fingerprint recognition technology based on embedded are of great significance, whether for fingerprint identification separating from the PC, or for the development of embedded systems applications.This thesis, first of all, made research of the characteristics of fingerprint recognition technology, applications and market. Discussed the development of a series of embedded fingerprint technical issues involved, given the general development of these embedded systems, making research of the Fingerprint collection instrument from theory and parameters. And given a comparison to several types of the Fingerprint collection instrument. Then, researched the fingerprint identification system based on ARM embedded. Analyzed The embedded fingerprint identification system From both hardware and software. Analyzed the likely problems of the design and development process. Gives the hardware design block diagram and design ideas, and analyzed the process of software algorithms. Finally, focused on the fingerprint recognition algorithm of the embedded systems. Made improve to the method of fingerprint image segmentation; improved the fingerprint image enhancement algorithm based on Gabor filter. Refining the use of two methods of refinement of the fingerprint image processing, to refine the use of traditional methods of fingerprint binarization of map size, and then use the improved algorithm OPTA thickness map of the second thinning. For each step in the process of the Fingerprint Identification System algorithms done a number of simplification and innovation, to make them more suitable for ARM-based embedded system running. All algorithms have been programmed in PC.