The Research And Implementation Of Embedded Fingerprint Identification Technical Based On FPGA

The fingerprint identification technology is one of biometric identification technologies, and the uniqueness and stability so that it has incomparable superiority of the other means. With the advancement of technology and social development, acquisition instrument and chip cost-effective continuous improvement; fingerprint recognition algorithms are becoming increasingly sophisticated; he security requirements of mobile devices continues to increase. Application of the fingerprint recognition technology in all walks of life:from the initial administration of justice, finance, etc. areas of expertise to the PC, mobile phones and other personal electronic consumer products. At present, fingerprint recognition device in electronic products is much base on DSP (Digital Signal Processor) technology. However, the DSP can only be executed sequentially and the identification system speed is limited. In contrast, FPGA (Field Programmable Gate Array) owns features with parallel processing, so designing the fingerprint identification IP Core based on FPGA has important practical significance.We study the fingerprint recognition algorithm for embedded environments, and propose a fingerprint classification algorithm in this paper; Moreover, we design fingerprint recognition algorithm based on FPGA hardware circuit, and apply the results to the SOPC (System on the Programmable Chip) embedded systems. We complete embedded fingerprint identification system in the FPGA environment, and get good results. Our main work is listed as follows:1. Study fingerprint classification algorithm, and propose fingerprint classification method based on the fusion of membership. This method applies LBP as characteristics to divide the fingerprint pattern area into sub-block, count histogram vector, classify fingerprint with the fusion of membership. It achieves good classification results.2. Research fingerprint recognition algorithm IP core in Verilog HDL (hardware description language). We build a simulation platform based on Matlab and the System Generator. We select top-down design approach, divide dedicated hardware circuit module into pre-processing, feature extraction and fingerprint matching by function, and optimize the design, ultimately achieve the fingerprint recognition algorithm IP Core. Compared to software implementation of the algorithm, each module’s performance has been greatly improved.3. Hardware circuit of sensor’s interface is designed on the FPGA to control the fingerprint sensor. We add the method of fingerprint image quality evaluation to the sensor control module. It ensures that the quality of fingerprint images input is high, and the performance of fingerprint recognition improved greatly.4. Fingerprint recognition algorithm IP Core and fingerprint classification algorithm are integrated into embedded systems. MacroBlaze soft-core CPU and Standalone software platform constitute an embedded system, and we achieve the FPGA embedded fingerprint identification system based on IP Core.The experimental results show that the fingerprint recognition algorithm IP core implement in this article is efficient, stable. Modular design is good for upgrade and secondary development, and IP Core can be quickly applied to different application environments. In addition, the FPGA embedded fingerprint identification system based on IP Core has unique advantage in terms of speed, power and expansibility, with broad prospects for development.