Automated Fingerprint Identification System Based On Dsp And Fpga Hardware Design And Realization

With the development of electronic computer and information technology, the biometric identification technology has drawn more attentions. Fingerprint identification technology as a part of biometric technology is developing very fast and used in many fields.Embedded fingerprint identification system is the one of the fingerprint identification applications. In tranditional embedded fingerprint identification systems, a single DSP or MIPS processor is mostly used to do the arithmetic. Because those processors work sequently, they could only match fingers one by one. When coming across a large finger lib, this processing would spend a lot of time. To solve this problem, a FPGA is added to associate DSP to do the matching arithmetic. As the FPGA can do matching processing in parallel and DSP has more accuracy, this new construction has high efficiency.The work about this thesis is gived below:We designed a hardware system including DSPs, FPGA, fingerprint sensor, human-machine interface and USB1.1 port. And we also wrote the related drivers to support applications. Because the DSP work at 300MHz, and some other chips work at 100MHz, in this thesis we discussed some signal integrity theories gived some experiences about PCB design.We made a FPGA design that 9 lines of finger match processing can be done at the same time. To avoid FPGA's disadvantage, we simplized finger match theory. And we also designed an I/O module to communicate with DSP.We designed an application program to identify fingerprint automaticly. And we also designed a structure to manage fingerprint database.Finally, this system is reliable, and it can identify fingerprint with high efficiency. And the modularized hardware and software are fitable in many different situations.