An Improved De-noising Algorithm based on Empirical Mode Decomposition for Respiratory Sound Signals

An enhanced noise-reduction algorithm utilizing empirical mode decomposition (EMD) has been introduced for the analysis of respiratory sound signals. As a method primarily driven by data, EMD offers significant capabilities in filtering out unwanted noise from signals. Combined with the wavelet thresholding principle and Hermite Interpolating Polynomial. In this study, we present a method that effectively distinguishes and separates the signal from the noise. Through the application of a defined threshold, we can acquire the modulus maxima for each intrinsic mode function (IMF) resulting from EMD decomposition. We employ the piecewise quintic Hermite Interpolating Polynomial to rebuild these IMFs. Upon successful reconstruction, the cleaned signal is derived by linearly superimposing the IMFs. Some experiments and simulation are carried out to confirm the efficiency of the improved de-noising algorithm. The results show that the result of classification is better using the improved EMD de-noising method than the result without using the de-noising method.