论文部分内容阅读
针对宽带噪声的特点，探讨了利用谱减法的改进形式进行语音增强处理，以实验结果说明了参数ｍ、ｎ的取值与增强效果的关系，当ｍ、ｎ取值恰当，则可获得较为理想的增强效果．在估计噪声的短时能量谱时，采用了低信噪比自适应有声／无声算法来判别纯噪声的信号帧，实时估计噪声，提高了噪声估计的准确度，取得了较好的语音增强效果．
Aiming at the characteristics of wideband noise, the improved form of spectral subtraction is used to enhance the speech. The experimental results show the relationship between the values of m and n and the enhancement effect. When the values of m and n are appropriate, Enhancement effect. In estimating the short-time energy spectrum of noise, a low signal-to-noise ratio adaptive speech / speechless algorithm is used to distinguish the signal frame of pure noise, to estimate the noise in real time and to improve the accuracy of the noise estimation and to obtain a better speech enhancement effect .