Research On Mobile Identification System Base On Biometrics

With the increasing requirements for security, biometrics based personal identification methods have received extensive attention.Recently, image pre-processing and embedded personal identification system are becoming an active topic in biometrics.This thesis takes a research on the image pre-processing algorithm and discusses the plamprint identification system based on embedded system.The main contributions of this thesis are summarized as follows:1,The diffcult problems in image pre-processing algorithm of the plamprint and iris identification system are discussed.A new algorithm based on inertia moment is proposed to locate the palm.First,calculate the orientation of the palm using inertia moment in the binarization image;then locate the palm through morphological filter.A novel iris coarse localization algorithm is proposed to assist the subsequent precise localization. First, a spot-search kernel is developed to locate all the light spots in the iris image;then the approximate center of the pupil is estimated.2,Introduce the idea of ordinal feature which is using in the state-of-the-art iris recogniton system. The algorithm is implemented on the palmprint recognition system which is proposed by Dr Zhenan Sun.3,A palmprint verification system based on ARM9 embedded system is designed.By virtue of embedded Linux transplanted to ARM, device drivers and applications for USB camera module and palmprint image acquisition program are developed.