基于DSP的嵌入式语音识别系统的研究与实现  

Embedded Speech Recognition System Based on DSP

作　　者：贾克明[1] 陶洪久[1] 

机构地区：[1]武汉理工大学信息工程学院,湖北武汉430070

出　　处：《武汉理工大学学报（信息与管理工程版）》2006年第7期156-159,共4页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：教育部重点实验室开发基金资助项目(TKLJ0208)

摘　　要：设计并实现了一个特定人、孤立词和小词汇量的嵌入式语音识别系统。系统硬件的核心芯片采用普遍使用的16位定点DSP芯片TMS320VC5416,软件上主要采用动态时间规整算法来实现语音识别。给出了系统整体的软硬件框架,并比较和分析了分别将线性预测倒谱参数和美尔频标倒谱参数作为语音特征参数时系统的性能,为语音识别的嵌入式应用提供了参考依据。A speaker dependent,isolated word and small vocabulary embedded speech recognition system is designed and implemented.TMS320VC5416 16-bit fixed-point DSP is adopted in its core processor of the hardware.Dynamic Time Warping(DTW) algorithm is chiefly used to implement the software.The hardware and software architectures of the system are presented.The performances of the system are compared and analyzed when the feature extraction is based on Linear Prediction Cepstrum Coefficient(LPCC) or Mel Frequency Cepstrum Coefficient(MFCC).The system is very useful for the embedded application of speech recognition system.

关 键 词：嵌入式 语音识别 DSP 线性预测倒谱参数 美尔频标倒谱参数 

分 类 号：TN912.34[电子电信—通信与信息系统]