论文部分内容阅读
提出一种基于匹配子波变换的罗音检测及分类方法。我们首先建立罗音信号的数学模型，然后基于这一模型设计子波变换的母函数。原始信号经过连续子波变换后，采用“软门限”进一步滤除噪声。根据每一尺度信号的能量，我们可以确定针对该信号的最佳尺度，从而可以检测罗音及对其进行分类。文中给出了实验方法及实验结果。
A method of rales detection and classification based on matched wavelet transform is proposed. We first establish the mathematical model of raleigh signal, and then design the mother function of wavelet transform based on this model. The original signal after continuous wavelet transform, the use of “soft threshold” to further filter out noise. Based on the energy of each scale signal, we can determine the best scale for the signal so that it can be detected and categorized. The experimental methods and experimental results are given in this paper.