Design Of Finger Vein Recognition System Based On ARM And WinCE

Along with the arriving times of information, people pay high attention to information security and secrecy. Fingerprint recognition is considered as a comparatively mature technology in biological feature recognition, which has well applied in many aspects of our life such as safety fixture, examining system, entrance guard, and so on. However, fingerprint features and information are exposed to the surface of body, which wil be subject to geographical and environmental impact of the outside world, making the identification of identity and identification accuracy declined.Finger vein identification technology relies on human finger vein feature and it has tremendous potential applications as the second generation of biometric authentication techniques, and it attracts many scholars. In order to achieve the small size of equipment, aiming at independence of the embedded system, we study the combination system of finger vein identification technology and embedded system, and the embedded finger vein recognition system based on ARM and Windows CE is designed and implemented. And it is more suitable for many cutting-edge occasions.This dissertation has done some elementary research on embedded finger vein recognition system, and finished the corresponding application development and the complete finger vein recognition system. Several main aspects included in this dissertation are as follows.For the part of hardware, the Intel Samsung S3C6410 processor and its peripheral devices (including image acquisition modules, interactive modules, etc.) are constituted the embedded hardware system. And the Windows CE operating system is builded and the relevant driver is designed, and successfully transplanted to the ARM11 board, and finally the whole embedded system platform is finished.For the identification algorithm, the process including format conversion and graying, gray normalization, image orientation and border cutting, vein division, filtering and denoising area, feature extraction and matching have been used. The effective methods for every steps of the process are analyzed in detail, and the characteristics of available algorithms are expounded. The suitable algorithm for every step according to the acquirement of the embedded system is selected, and the related development according to situation itself is finished by hardware structure. The paper describes detailedly the thinning algorithm of intravenous lines, and the feature points is extract through eight neighborhood encoding detection method. Considering the complexity of the algorithm and other factors to affect the operation system, the feature matching method using the MHD algorithm is used, and it is conformance to requirements through the experimental results.Meanwhile, for the part of software, the finger vein image acquisition and recognition application development is completed by using of the multimedia technology. At last, the embedded identification system is tested, and good experimental result is achieved.