基于DSP的语音识别智能控制系统

Controlling of brainpower of speech recognition system based on DSP

介绍了语音识别的基本原理及用浮点数字信号处理器(DSP)TMS320C32实现语音识别算法的一些原则和方法,阐述了语音识别的DSP实现技术,系统以预测倒谱参数为特征参数,并采用计算量相对较小的改进的动态时间规整(DTW)算法实现语音参数模板匹配,能够实现特定人、孤立词、小词汇量的语音识别,并用MATLAB进行了算法仿真,从而将语音识别技术应用到智能控制系统中,给出了实验结果和误差分析。试验结果表明,系统正确识别率在89.96%,具有一定的实用价值。

This paper introduces the fundamental theory of speech recognition, and some principles and ways on implementation of it in float-point TMS320C32, explain the DSP realization technology of speech recognition. The system of speech recognition adopts linear prediction cep strum coefficient （LPCC） to catch speech characteristic parameters and introduces dynamic time wrapping （DTW） arithmetic to realize speech pattern matching. It is proved that this article designs a small vocabulary, given human, isolated word speech recognition system, arithmetic of Speech recognize simulate with MATLAB software, and applies them into controlling of brainpower. It it proved in experiment that systematic correct recognition rate reach 92%, and has some realistic application value.