Embedded Fingerprint Recognition System Based On ARM9

Fingerprint identification technology aims to achieve personal identification by the computer using thefingerprint of human-specific physiological characteristics. People’s fingerprint is unique and can not becopied, and do not have to worry about to be lost, so it is called evidence of the first, as the most widelyused biometric identification method in the various methods of human biometric identification, fingerprintidentification has a very high security and easy to use. Automatic fingerprint identification technologyincluding three stages, fingerprint image preprocessing (fingerprint segmentation, fingerprint enhancement,fingerprint thinning), feature extraction, feature matching. Early fingerprint recognition system needs to beprocessed on a computer, the portability of the system is poor, and that limits its application. With theevolving and progress of embedded technology, and the performance of ARM embedded processor is rising,that makes embedded fingerprint identification technology to practical use.In this paper, embedded fingerprint identification technology being studied firstly, the characteristicsof embedded fingerprint identification technology being introduced, discuss technical issues related toembedded fingerprint development and practical application, discuss the basic development steps ofembedded fingerprint identification system, analysis the applications, markets and trends.Then in embedded Linux environment, from hardware and software to start building a fingerprintidentification system. On the hardware side, the more popular Samsung ARMS3C2440microprocessor isused as the core processor, on the software side, fingerprint image acquisition, image preprocessing,fingerprint feature extraction and ant algorithm for feature matching method are discussed. Finally, theexperiment results show that the identify system has the characteristics of higher recognition accuracy,strong anti-interference ability, and shorter match time. In this paper the transplantation of embeddedfingerprint recognition algorithm is verificated by experiments. This study of this paper is lain thefoundation for the embedded fingerprint identification system, the embedded fingerprint identificationtechnology can be used widely.