Research And Implementation On Embedded Iris Recognition System Based On DSP

Iris recognition gets more and more attention in recent years. Compared with the other biologicalcharacteristics, iris has many outstanding characteristics, such as uniqueness, universality, stabilityand security. Iris recognition has been widely used in many fields, such as daily attendance, airportsecurity and national security. With the high recognition rate and security, iris recognition has becomethe research focus in the field of biometrics recognition. Because of the big volume, high cost andmobile inconvenience, embedded iris recognition system based on DSP is proposed in this paper toprovide theoretical and technical support for the development of small, convenient and offline irisrecognition system.Based on the research of the iris recognition key technology and DSP basic principle, the irisrecognition original system is designed according to the iris recognition performance requirements.The main reaearch contents of this paper are as follows:(1) Based on the research of the iris recognition key technology and DSP basic principle, suitableDSP is selected as the kernel of the embedded iris recognition system.(2) The principle of designing embedded hardware system has been studied. According to thecharacteristics of isir recognition system, the embedded iris recognition hardware system based onDSP is designed, namely, power management, image acquisition, image processing, data storage,human-computer interaction and network communication modules for detailed design.(3) On the study of embedded operating system DSP/BIOS, the embedded iris recognitionsoftware system is developed by the DSP integrated development environment (CCS). According toiris recognition function modules, the tasks are partitioned and the embedded iris recognitionalgorithm is realized. Based on the research of optimization method of embedded system performance,the performance of the embedded iris recognition system is optimized in order to achieve theperformance requirements or iris recognition.The performance of embedded iris recognition system has been tested with multiple iris imagelibraries. The results show that the embedded iris recognition system is fast and reliable with highrecognition rate.