Research Of Dual-mode Embedded System Based On Fingerprint And Finger Vein

With information security increasingly taken seriously, personal identificationrecognition technology which based on biometrics gains extensive attention. At present, thepersonal identification recognition products which available in the market are mainlyembedded systems of unimodal biometric. However, a unimodal biometrics technologyusually was far from perfect in many real applications, and it is difficult to achievebreakthrough results of algorithm improvement and acquisition equipment ascension aimingfor one biometrics technology. This problem can be allivated by the use of multiple sensorscapturing biological characteristics, extraction and fusion of different kinds of biologicalcharacteristics image in the different stages, which forms multi-modal biometrics system. Themajority of the multi-modal biometrics technology research have so far been tried only in thelaboratory. This thesis carries out the research of the embedded multi-modal biometricssystems, drives the personal identification recognition technology which is more effective,accurate and broader based on the platform of embedded system to the market.This paper makes a design of the embedded multi-modal biometrics system based onARM-WinCE6.0, which primarily focuses on the following two parts: the hardware platformand the software platform. The hardware platform principally includes the ARM coremotherboard controlling the whole system and operating the related algorithm, the systemexpansion board realizing acquisition and display and providing an interface to the relatedperipheral equipments, the image acquisition board making a collection of the multi-modalimage. The software platform mainly includes the design of the bottom driver and the upperapplication software.Firstly, in the design of the hardware platform it analyses the functions that the hardwareshould achieve, selects the corresponding components and gives the related circuit diagramsor principle diagrams. The design of the image acquisition module is introduced chiefly andparts of the debugging process are given. Then in the design of the software, it focuses onintroducing the architecture of the WinCE6.0system and the stream driver. On the basis of thestream driver, we design and realize the driver for the OV2640camera with the combinationof the DirectShow characteristics in the development environment of WinCE6.0, cut andcompile the whole system, make transplantation afterwards. Thirdly, combining with thefunctions of the embedded system’s modules in the image acquisition and display,theapplication is designed and implemented, compiled and transplanted into the embedded system in the development environment of VS2005. At last, the embedded system isdebugged, some phenomenon of the software and the hardware is analyzed, the debuggingmethod of the key steps is given. The experimental results demonstrate the feasibility of theembedded system, creating a good platform for later algorithm research of the multi-modalbiometrics technology.