| With the rapid development of society,the needs of identification of individuals growing andeverywhere.The performance of traditional identification methods limitations and drawbacksbecome increasingly apparent.The biometric identification technology which emerges quietly canessentially solve the traditional identification of technical deficiencies.It has gradually beingrecognized.Fingerprint identification has become the most reliable and user-friendly type ofbiometric identification technology.At present,the traditional PC.based fingerprint identificationsystem has many shortcomings, for example,large, not easy to carry, poor real-time and so on.Thusthe fingerprint identification system which based on the embedded has become a hot researchnow.In this context,fingerprint recognition technology based on embedded are of greatsignificance,whether for fingerprint identification separating from the PC,or for the development ofembedded 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 technicalissues involved,given the general development of these embedded systems,making research of theFingerprint collection instrument from theory and parameters. And given a comparison to severaltypes of the Fingerprint collection instrument.Then researched the fingerprint identification systembased on ARM embedded.Analyzed the embedded fingerprint identification system from bothhardware and software.Analyzed the likely problems of the design and development process.Givesthe hardware design block diagram and design ideas,and analyzed the process of softwarealgorithms.Finally,focused on the fingerprint recognition algorithm of the embedded systems.Madeimprove to the method of fingerprint image segmentation;improved the fingerprint imageenhancement algorithm based on Gabor filter.Refining the use of two methods of refinement of thefingerprint image processing,to refine the use of traditional methods of fingerprint binarization ofmap size.For each step in the process of the Fingerprint Identification System algorithms done anumber of simplification and innovation to make them more suitable for ARM-based embeddedsystem running.All algorithms have been programmed in PC. |
