The Research Of Key Technologies On The Embedded Safety Verification System Of Online Payment

Currently, USB-Key uses the traditional password authentication method, which has the imperfection of easy lost and forgetfulness. The imperfection gives rise to a certain lack of security and convenience. In this situation, a USB-Key utilizing fingerprint image as the payment code was designed to solve the disadvantages of traditional USB-Key effectively.Three various processor platforms, i.e. S3C6410, Xilinx Zynq and STM32F429, were used for system design, such as fingerprint image acquisition, processing and recognition. The key technologies of fingerprint recognition were analyzed in detail, and the underlying interfaces and startup codes of different platforms were designed and compared. Meanwhile, the fingerprint recognition algorithm was successfully transplanted on the three embedded systems respectively. Eventually, the acquisition and recognition performances of three experiments were compared.The fingerprint recognition system mainly consists of three parts, i.e. CPU underlying interface and startup code design, USB transport interface design and fingerprint image algorithm design. On the FPGA platform based on Xilinx Zynq processor, fingerprint image acquisition, image direction characteristics extraction and SDRAM data caching interface were realized by HDL code design and the timing simulation were conducted. According to the test and verification, the fingerprint recognition time cost on the platform based on processors S3C6410, Xilinx Zynq and STM32F429 were 7.8s, 0.27 s and 1.3s respectively.