| The work of this paper focuses on making the whole speaker recognition system that can recognize a speaker in the real-time. This system has been used in a phonetic electronic lock in the floor door elementarily. It also researches the development of Biologic Recognition Technology and speaker recognition algorithm based on vector quantization.Biologic Recognition Technology definiens the automatic recognition technology based on human circadian and behaved character through computer. It has been proved that the fingerprint, palmprint, sound, iris, retina and framework are exclusive and steady which can be used to recognize who are you. Using these characters, we have palmmprint recongnition, fingerprint recongnition, face recongnition, speaker recognition, iris recongnition, idiograph recongnition and so on. By far, the domain based on Biologic Recognition Technology in USA is some hundred millions dollar and other countries alse use Biologic Recognition Technology by the domain of law. On the whole, Biologic Recognition Technology is used in the actual application in government and business, the main application fields include passenger controlling in airdrome, the government, door system, checking on work attendance, law administer, consumer controlling, finance controlling, computer landing, medical treatment and health protection so on.The work of this paper focuses on making the whole speaker recognition system that is the application of Biologic Recognition Technology in the joining door. The voice is exclusive so that we can recongnize speaker exactly. By compare with existent door systems, it is cheaper, more convenient, more maintainable and more healthful. So the system has more capacious market potential.The main work and content are sum-up as follow:The dissertation consists of six chapters.In Chapter 1, it explains the history and actuality of Biologic Recognition Technology and introduces the advantage of speaker recongnition used in door. I also put forward the research and the main work of this paper.In Chapter 2, researching on the main methods of speaker recognition, I focus on speaker recognition algorithm based on vector quantization. Finally, this paper gives the demand of hardware and software by the recognition algorithm.In Chapter 3, designing the controller of the spesker recongnition system and devising the algorithm cell using the core of TMS320VC5402 DSP.In Chapter 4, designing the administrator of the speaker recognition system, which can be used to manage the system by man.In Chapter 5, designing the correlative software of the speaker recognition system and bringing forward a new communication design of DSP and PC.In Chapter 6, a brief summary is set out. And the research direction and future work are given.The main contribution of this dissertation is listed as follow:â… Collecting and cleaning up a lot of literature about Biologic Recognition Technology . By comparing all kinds of Biologic Recognition Technologies, the speaker rocongnition system has more capacious market potential.â…¡Designing the contoroller and administer of speaker recongnition system and finishing the hardware design of speaker recognition algorithm based on vector quantization. At the same time, the detached design meet the need of natural using and centralized administer.â…¢ Designing the correlative software of the speaker recognition system, realizing the RFFT in DSP and bringing forward a new communication design of DSP and PC.â…£ Finishing the speaker recognition system which is the experimental equipment of phonetic electronic lock in the floor door. And the design is an all-right reference for other DSP system designs.And this paper has some heuristic value on the two questions as follows.â… How to widely study the more suitable speaker recognition algorithm for the application of DSP implementation.â…¡How to consummate the hardware system design and optimize the software design based on speaker recognition algorithm, in order to enhance the whole capability.
|
PDF Full Text Request