Design And Implecation Of Voice Print Recognition System Based On SOPC

The era of "unsupervised" identity authentication has arrived.The rapid,accurate and convenient realization of identity authentication can not only facilitate people's lives,but also protect personal rights and interests.The risk of "impersonation" is widespread in the past common identity authentication methods.Therefore,biometric identification technology arises at the historic moment.Voice as a natural attribute of human carry,the use of "contactless" authentication method,make the identity authentication more convenient,and the hardware requirements are simple,in the criminal investigation,finance,military,intelligent equipment and other fields,can be widely used.SOPC(System-on-a-Programmable-Chip)is a system based on FPGA(Field-Programmable Gate Array),which is realized through the combination of hardware and software.It has the advantages of safety,high efficiency,low cost and easy maintenance.Therefore,this system is designed and implemented based on SOPC technology.This paper designs and implements a voice print recognition system based on SOPC.The main content includes the design and realization of hardware module based on FPGA and the integration of each module into SOPC system.The hardware part includes two modules: preprocessing and feature extraction.The preprocessing module includes the preweighting module,the framing module,the windoring module and the endpoint detection module,which realizes the preliminary processing of the speech signal,eliminates the noise and enhances the useful information.The feature extraction module includes FFT module,power spectrum module,MEL filter module,LOG module and DCT module,which can extract the characteristics that can express the speaker's unique personality from the preprocessed speech signal.Combined with Nios II processor,SOPC system is designed and integrated.The test results show that the system is effective and feasible.