Wavelet analysis of electrocardiogram signal
GU Zhen-pu · Hebei ke-ji daxue xuebao · 2006
The electrocardiogram signal is a unstationary weak signal with a lot of distinguished vertex.The noise elimination method of the mold maximum value in the wavelet transformation has nonlinearity and the quality of self adaptation.The quality of the wavelet is very applicable to the unstationary weak signal which resembles the electrocardiogram signal.In this paper, the time to frequency localization quality of the wavelet,and the signal wavelet resolution and reconstruction algorithm Mallat which is based on the multiple resolution analysis,were studied according to the limitation of the traditional noise elimination method in the processing of the electrocardiogram signal.The detection of the R peak value of the QRS wave and the noise elimination were carried out by the method of the mold maximum value.The experiment indicates that the identification rate of the QRS wave is to be 99.9% through the detection of the wavelet.The signal-to-noise ratio of the electrocardiogram signal has been greatly improved on the primordial signal by the reconstruction of the wavelet algorithm.