Design Of Embedded Fingerprint Acquisition System And Preprocessing Algorithm Research

Biometrics is the development direction of the future identification technology. Automatic fingerprint identification is one of the most reliable and effective biometric technologies, with the development of IC technology, Automatic fingerprint identification system is gradually moving away from the computer in flexibility, convenience and real-time directions. Research and development of the small, intelligent acquisition and identification system is the fingerprint recognition technology to practical research.This thesis design and implement an embedded fingerprint collection system by using TMS320VC5509DSP and MBF200 as mainly devices, SPI as the communication interface. First, the thesis deplly study the structural characteristics of DSP,MBF200 and their peripheral devices on the basis of design the overall system structure, finish hardware circuit of fingerprint acquisition module and algorithm processing system. Second, development fingerprint acquisition program under CCS2.0, obtain better image and analysis every parameters of effect fingerprint quality. Finally, according to characteristics of acquisition fingerprint, select suitable filters for pattern filtering enhanced on the basis of segmentation and normalization, select dynamic threshold method through comparison various binary algorithms, introducing mathematics morphology for thinning treatment, design one set of pretreatment algorithm.The thesis completed fingerprint acquisition and pretreatment. Size of fingerprint image is 152x200 pixel, reading of fingerprint sensor can reache up to 12Mb/s, dermal ridge is clear, the effect of thinning is significantly after pretreatment. futher popularization of the fingerprint technology is enhanced.