The Research And Design Of Fingerprint Identification System Based On DSP

With the development of society, people put forward higher requirements for the security and convenience of identity authentication. In this context, the use of biometric technology to authenticate the identity has been more and more widely used. Compared with other biometric technologies, fingerprint identification technology has the advantages of unique, large amount of information, easy to identify errors, easy to collect, easy to accept, and so on. Due to the continuous progress of chip technology and fingerprint identification technology, the development of a high performance fingerprint recognition system with high performance is of great significance and market value. In this paper,using TI's TMS320C6748 DSP designed a complete set of fingerprint identification system, the system can on-line debugging and running, also can be stand-alone operation. The system can realize fingerprint acquisition,image display and human-machine interaction and so on.First of all, this paper design the overall scheme of fingerprint identification system and finally decided to adopt TMS320C6748 DSP of TI company as the main processor. The main processor's external extended DDR2 SDRAM and NOR flash memory device, fingerprint sensor interface, LCD and touch screen human-machine interface module.According to the system scheme,selecting chip model of every module. And refer to the chip user manual datasheet, completed the circuit schematic design of each module on the Capture CIS OrCAD software, and design the driver program for the main chip. Complete the design PCB layout, wiring in the Allegro PCB software. Take modern high-speed PCB design method when to deal with high-speed DDR2 SDRAM, strictly in accordance with design specifications of DDR2 SDRAM, and do pre simulation and post simulation analysis, to ensure that the design is correct.After the completion of the design of the hardware platform,this paper analyses the fingerprint recognition algorithm of pre-processing, feature points extraction and feature points matching process in detail. Before the algorithm is transplanted into DSP, simulated the preprocess of fingerprint image including smoothing, segmentation, enhancement, thinning and binaryzation on the platform of Visual Stdio and achieved the expectation of fingerprint image processing effect. Then the extraction and matching of feature points are analyzed respectively.Test the minimum system, SDRAM DDR2, fingerprint acquisition, humanmachine interaction and other modules respectively after proofing and welding the circuit board,test results show that the design of the hardware circuit is correct. After the completion of the test of the hardware platform, loading fingerprint recognition algorithm into the DSP to simulate and debug. The processing effect of fingerprint recognition algorithm on DSP is the same as the Visual Stdio platform. At this point,the design tasks of fingerprint identification system have been completed basically.