Research And Implementation Of ECG Signal Denoising And Recognition Based On Wavelet Transform

Heart disease is one of the most serious diseases that threaten human life. People have never stopped studying the disease. With the increase of the number of patients, the automatic analysis of ECG will become the trend of the future. The automatic diagnosis of heart disease is based on the characteristics of ECG wavefonn and some information, which is based on pure ECG signal. Therefore, in order to promote the development of the technology of automatic analysis of ECG signal, to accelerate the pace of heart disease prevention, diagnosis and treatment, this paper focuses on the automatic analysis of ECG signals preprocessing and waveform detection technology are studied. The main research contents are as follows:1. Preprocessing and waveform detection algorithm of ECG signal is proposed in this paper using data developed by the laboratory of ECG signal acquisition device to collect the ECG data and MIT/BIH arrhythmia database in the verification.2. The preprocessing of ECG signal is studied. Based on the stationary wavelet transform and threshold denoising method, three kinds of noises,such as baseline wander, power frequency interference and EMG interference, are eliminated. First of all, to make a comparative analysis of commonly used wavelet functions, stationary wavelet selected wavelet function bior5.5 8 scale on the ECG signal decomposition; secondly, in order to overcome the defect signal amplitude appear soft and hard thresholding function in signal reconstruction distortion, edge blur and pseudo Gibbs phenomenon, a new threshold function is proposed, and threshold denoising based on threshold heursure. The experimental results show that the proposed method can effectively remove the three kinds of main noise in ECG signal, and obtain a high signal to noise ratio.3. The study of ECG waveform detection. By analyzing the first derivative of the smoothing function, which is the two B spline wavelet as the wavelet function, the ECG signal is decomposed into 4 layers. The detection of R wave is realized by using the modulus maxima generated by the R wave on the fourth scale detail coefficient cd4. The detail coefficients of CD2 second standard, in order to detect the R wave generated by the maxima of the basis, in the left and right sides by detecting extremum of zero, so as to realize the detection of Q wave, S wave and QRS wave starting point and end point. According to the Q wave, R wave, S wave and the starting point and end point of QRS wave detection and puts forward the corresponding prevent and prevent the error detection and position correction mechanism, to further improve the accuracy of waveform detection.