Design And Implementation Of Dsp-based Iris Recognition System

In high information modern society, people need to use the identity authentication in daily life. Iris recognition technology is a potential recognition technology in the current status of identity authentication. Compares with other recognition technology, the iris recognition has many other advantages, such as the stability, uniqueness, can be collected, non-contact and so on. So the iris recognition has the broad application space and the good economic efficiency, attracts domestic and foreign scholars more and more attention.At present, the embedded iris recognition systems are mostly based on DSP + FPGA system architecture, but this system use only one DSP core (TMS320DM642), which can complete image acquisition, image processing, and communication with the peripheral circuits. The whole system not only can reduce costs, shorten the development cycle, and also can reduce power consumption.This paper mainly introduced the development of the embedded iris recognition system. The work is composed of the following several parts.Iris recognition's hardware team has completed the architecture of embedded iris recognition system, including the DSP processor module, power modules, memory modules, video interface module and so on.The iris recognition hardware team together debugs the whole system's hardware module, and develops the corresponding low-level driver modules.Completed the transplant of the embedded operating system UC/OS-II on TMS320DM642, it runs stably in the TMS320DM642.Completed the entire application development of iris recognition system, based on UC/OS-II and the low-level driver, transplanted the iris recognition algorithm to the system with C language, and optimized the program through many ways.Transplanted the open-source TCP/IP protocol stack LWIP in TMS320DM642. Completed a simple embedded web server, and it is able to communicate with the outside world stability.Finally, The paper complete the design of hardware system architecture, the development of low-level driver, the transplant of embedded operating systems and TCP/IP protocol stack, the development of system application and so on. Each module could run stably, the whole system can achieve an embedded iris recognition system's basic functions. Because the design of software and hardware follows the standard, it facilitates system upgrades or second development in future.