【摘要】 介绍了一个实时语音识别系统的设计与实现。该系统硬件以TI生产的TMS320C5409 DSP为核心,配置编解码器、外部存储器和逻辑控制电路等。软件算法采用了由能量和过零率相结合的端点检测、Mel频率倒谱系数特征提取和径向基函数神经网络构成的语音识别算法。对汉语数字孤立词进行了实时识别实验,识别结果通过异步通信串口送到PC上实时显示。实验结果证明了系统的有效性。更多还原

【Abstract】 This paper introduces design and implementation of a real time speech recognition system,which is based on TMS320C5409 DSP produced by TI the system′s hardware configure codec,external memory and logical controlled circuit etc.Speech endpoint detection of the average instantaneous energy and the average instantaneous zero crossing rate(ZCR),zero-crossing with peak amplitudes feature extraction and radial basis function neural network are used in the system.Validity is proved by real time recognition experiment of Chinese character digit isolated words.The experiment results are sent to PC through asynchronous corresponding serial port.更多还原