摘要
语音增强技术是语音信号处理的一个重要分支,也是语音识别系统的重要组成部分。简单介绍了现有的各种语音增强算法,并对目前应用最为普遍的谱相减法作了系统的描述。对传统型、增强型、改进型的谱相减法进行了探讨验证,提出了一种基于C语言环境的算法程序,详细地介绍了该程序并画出了它的算法流程图,而且进行了算法仿真,通过分析它在-5dB,0dB,5dB的信噪比下的输入输出波形,证明了该程序可以较好地实现语音增强。
Speech enhancement technology is an important branch of speech signal processing,but also the important component of speech recognition system.The existing algorithm of speech enhancement is introduced briefly,and applied the current application of the most popular spectral subtraction of the system.Discussed and verified for traditional,enhanced and improved the spectral subtraction,proposed the algorithm procedures based on C language environment,the detailed algorithm is introduced in this program and drew out the flow chart,through the analysis of the input and output waveform in-5 dB,0 dB,5 dB,proved the program can achieve speech enhancement.
出处
《科学技术与工程》
2010年第31期7652-7655,共4页
Science Technology and Engineering
