摘要
针对传统机械零件自动分类回收系统因复杂噪声环境导致语音识别准确性不高的问题,文章提出一种混合语音降噪算法,利用谱减法、维纳滤波与小波阈值降噪对语音进行多级降噪处理。搭建实验环境对改进后的方式进行验证,实验结果表明改进后的方式能够有效改善系统语音识别的准确性,提升系统分拣效率。
Aiming at the problem of low accuracy of speech recognition caused by the complex noise environment in the traditional automatic sorting and recycling system for mechanical parts,a hybrid speech noise reduction algorithm is proposed,which makes use of spectral subtraction,Wiener filtering and wavelet threshold noise reduction to carry out multilevel noise reduction for speech,and then finally builds up experimental environments to validate the improved method,and the experimental results show that the improved method can improve the system's accuracy of speech recognition and enhance the system's sorting efficiency effectively.
作者
于洪波
邵娟
YU Hongbo;SHAO Juan(Liaoning Construction Vocational College,Liaoyang 111000,China)
出处
《电声技术》
2024年第2期36-38,共3页
Audio Engineering
关键词
机械零件
自动分类
语音识别
语音降噪
mechanical parts
automatic classification
speech recognition
speech noise reduction
