摘要
本文基于当前嵌入式设备广泛应用的环境下,在语音识别技术的基础之上,设计了以ARM处理器为核心,Linux为操作系统的嵌入式语音识别设备。语音识别采用了流行的DHMM模型,并使用系统开销较小的Viterbi算法实现。总体来说,本文所设计的语音识别设备具有价格低、性能强、通用性好以及扩展能力强等优点。
This paper designs the embedded speech recognition equipment which have the ARM processor as the kernel and Linux as operating system based on speech recognition technology,in the environment where embedded equipments are widely used.Speech recognition selects popular DHMM model,and uses Viterbi algorithm which has a very small system costs.All in all,the system owns advantages of low price,strong performance,high universality and good extensibility.
出处
《价值工程》
2012年第4期126-127,共2页
Value Engineering
基金
云南省教育厅科学研究基金项目
项目编号:2010Y503
