Study On Voice Recognition System Based On Single - Chip Microcomputer Spce061a

Voiceprint Recognition is also called Speaker Recognition, which is a kind of speech recognition. Unlike speech recognition, voiceprint recognition distinguishes speaker from others by his or her voice, not the meaning by voice.Voiceprint Recognition can be applied widely, which has lots of successful applications from laboratory to practice in recent 50 years. Most of applications make use of computer as core processor; the speed and quantity of computer are very high, very excellent results have been got. But the disadvantages about computer are obvious. The price, volume and mobility have limited its applications. Many researches are devoted to the embedded system of voiceprint recognition.This article researched into the principle of voiceprint recognition and the methods of constructing a voiceprint recognition system. The key technologies about voiceprint recognition is analyzed. Finally, a proposal is put forward, which is based on SPCE061A to construct a speaker verification system for voiced speech. This proposal makes the best of the resources on the SPCE061A and extends a keyboard to show the results in hardware. The program in the singlechip computer, including endpoint-detection, Pre-emphasis, Linear Prediction Coefficient and Linear Prediction Cepstrum Coefficient, are designed on unSP IDE 1.16.1 by ANSI C and assembly language, which suited to operate on singlechip computer. According to the demands, another program is planed to convert binary files in singlechip computer into text files which can be analyzed by Matlab6.3.The testing results show that this system can complete some functions of recognition, but there are many aspects to improve. But it always has some significance on promoting the developments and applications on embedded voiceprint recognition.