Acquisition And Recognition System Based On Multispectral Palm

Any single modal biometric has its limitations, so as to affect the results and limit thescope of application. Combination of the hand-shape, palm-print, palm vein identificationtheory and DSP (TMS320DM642) digital processing system, Proposed and designed athree-modal fusion palm and palm prints and palm vein features embedded imageacquisition system. This mainly includes a multi-spectral light source subsystem, CMOSimage sensor, a high-frequency digital signal processor, a liquid crystal display module,and the voice prompt module.In order to improve the acquired modal image quality, this paper also focuses on lightsource subsystem spectrally selection, uniform analysis and logic control research, anddesignes an uniform illumination, programmable multi-spectral light source.470nm bluelight and850nm near-infrared light are the primary light source. AMC7150constantcurrent driver chip logic controls design illumination uniformity and programmablemulti-spectral light source.In order to ensure the system’s real-time performance and efficiency, Video Portinterface achieves the automatic acquisition of image data, and reduces the burden on theCPU, which can enable to focus on the identification of data processing work. CMOSdigital image sensor OV7620and VP2interface connectes seamlessly to complete thework of image data acquisition. Via the VP1interface the liquid crystal display moduleachieves the image display function. The voice prompts the module WT588D usesfirst-line control mode to achieve voice prompts and broadcast feature.The software system is designed on hardware platform, including image sensors,video display and voice modules initialization procedure, VP port acquisition and displaydriver and the light source control program to complete the framework of the entiresoftware system and related algorithms embedded.The system achieves real-time acquisition, processing and display images in theoffline situations. Not only does it meet the practical requirements, but also provides ahardware platform for the subsequent multi-modal characteristics of the imaging system design. The results indicating, the palm three modal acquisition system designed in this paper achieves real-time collection and meets the identification requirements of the palm-shaped, palmprint and palm vein images, and laying a good foundation for the next step in the data processing and recognition.