Fingerprint Recognition Based On Embedded Linux System And Implementation

As information-based rapid development of science and technology, has a higher security identification technology has become the first to consider the elements of many systems, based on the traditional password and token identification technology easily lost and stolen there are other hidden dangers, and through the the biological characteristics of human identity authentication provides a more reliable solution. Biometric technology is biometric-based, information processing technology as a means of biotechnology and information technology combine. Among the many biometric technologies, fingerprint recognition technology is the most widely used one, it has a universal, permanent, can be collected and easy realization of the advantages, has become the most effective biometric technologies. Fingerprint recognition technology will be combined with the embedded system can get rid of the fingerprint identification system and the environment of space limitations, more widely applicable to various occasions.With the computer and microelectronics technology, embedded fingerprint identification technology based on more and more people pay attention and study, embedded system has a processing speed faster and faster, richer variety of interfaces for a variety of electronic product features hardware and software development. Fingerprint recognition technology will be combined with the embedded system can get rid of the fingerprint identification system and the environment of space constraints, more extensive needs for all occasionsThis paper designs based on embedded linux fingerprint identification system, hardware platform architecture S3C2410 ARM9-based embedded test platform. Described in detail the structures of embedded Linux development platform, including the cross-development environment to build, Bootloader transplantation, Linux kernel reduction and transplantation, root file system, production and transplantation, the transplant-related device drivers, application development for the follow-up provides a good basis.In this paper, fingerprint recognition algorithm in-depth research, selection, and the improved algorithm is more suitable for embedded platforms, the algorithm is divided into three parts:fingerprint image preprocessing, feature extraction and matching. One part by the normalization preprocessing, segmentation, binarization, noise reduction, thinning algorithm composition, fingerprint matching algorithm is based on two-dimensional cluster of point pattern matching algorithm.Finally MiniGUI fingerprint identification system designed interactive interface, and the development of complete embedded Linux-based fingerprint identification system. Experimental results show that the stability of the system can run on ARM platform, the fingerprint recognition technology has great applications in the embedded reference platform.