摘要
系统以16位数字信号处理器TMS320VC5502为核心,采用音频Codec芯片TLV320AIC23对语音信号进行采集和编码转换,通过端点检测、特征参数提取、DTW算法等关键技术实现特定人、小词汇量、孤立词的语音识别,最终根据LED闪灯次数检测数字0～9的识别结果。
The system based on the 16-bit digital signal processor TMS320VC5502, using audio codec TLV320AIC23 to acquire and transcoding the voice signal, achieving a speaker dependent and small vocabulary isolated word speech recognition by endpoint detection, characteristic parameters extraction, DTW algorithm and other key technologies, ultimately according to the flash frequency of LED to detect recognition results from numbers 0 to 9.
出处
《科技广场》
2011年第7期118-122,共5页
Science Mosaic
