经验模态分解在咳嗽音检测中的应用

Application of empirical mode ecomposition in the detection of cough sound

咳嗽是众多呼吸道疾病中常见的重要病症之一,具有极其重要的临床信息。本文提出了一种基于经验模态分解(EMD)和Teager能量算子的咳嗽音识别特征,与传统方法提取的美尔倒谱参数相比,EMD方法的特征参数在个数上有大幅度的下降。试验证明,EMD方法的特征参数在高信噪比环境下的咳嗽音检测中是行之有效的。

Cough is a common symptom of many respiratory diseases. The evaluation to its intensity and frequency of occurrence provides valuable clinical information in the assessment of patients with cough. Empirical mode decomposition （EMD） and Teager energy operators（TEO） were applied to the characteristic parameters procurement in the detection of cough sound. Compared with the traditional method via empirical mode decomposition the dimension of the feature is reduced. Experiment results show that the new feature with EMD and TEO is efficient for detecting cough sound in high.