%0 Journal Article
%T 基于功率谱密度的打呼噜声提取研究<br>Power Spectrum Density-Based Snore Sound Extraction Research
%A 孙树平
%A 庞宏祥
%A 黄婷婷
%A 张弼强
%A 李肖航
%A 王楠
%J Open Journal of Acoustics and Vibration
%P 26-32
%@ 2328-0522
%D 2020
%I Hans Publishing
%R 10.12677/OJAV.2020.81004
%X <div style=\"text-align:justify;\">
	<span style=\"font-size:14px;\">为确定鼾声的特征，定义并提取了基于频率域的功率谱密度的鼾声特征，对正常和异常鼾声进行了评价。本研究分为三个阶段：首先，通过微信号采集和预处理，利用小波分解方法去除噪声。其次，利用功率谱密度函数(PSD)结合阈值法提取频域特征。最后，对十三名健康学生的三十二小时正常呼吸声和三名患者的十一小时异常鼾声进行了研究。分析结果表明，异常呼吸声对应的两个主成分分别分布在10~300 Hz和500~800 Hz范围内，比从正常呼吸声中提取的主成分高。<br />
To character the characteristics of snore sounds, the power spectrum density (PSD)-based snore sound features in the frequency-domain are defined and extracted to discriminate the normal and abnormal snore sounds. This study is generally divided into three stages: firstly, the snoring signal is collected via micro-recorder and is preprocessed to denoise the unexpected noise via wavelet decomposition method. Secondly, PSD combined with threshold line is employed to extract features in the frequency-domain. Finally, thirty-two hours normal breathing sounds from thirteen health students and eleven hours abnormal sounds from three patients are analyzed. The analysis results show that the two principal components corresponding to the abnormal sounds, greater than those extracted from normal sounds, are distributed in the range of 10~300 Hz and 500~800 Hz, respectively.</span> 
</div>
%K 呼噜声信号，小波变换，功率谱密度函数，呼噜声成分<br>Snore Signal
%K Wavelet Transform
%K Power Spectral Density Function
%K Snoring Components
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=34585