Arithmatic Research On Preprocess And Characteristic Recognition Of ECG

Electrocardiograph (ECG) signal is synthetical reflection of Cardiac Activity in body surface. It is important significance for heart to recognize. ECG processing and recognition is analyzed and in this paper. Based on those, researches of ECG preprocessing and recognition are implemented, some good effects Will be obtainedPower line interference and baseline wander maximal are interference to ECG diagnosis analysis, Signal-to-Noise of ECG is reduced enormously. Based on self-adaptive noise cancellation system, improvement arithmetic for least mean square (LMS) is introduced; LMS arithmetic for step change LMS Arithmetic based iterative time is bring forward. Then, characteristic wave of ECG is processed by wavelet based on multi-resolution. Aimed at the problem of error test for high amplitude pseudo difference to wavelet, integration Arithmetic of wavelet and morphological for QRS wave test is bought. This way doesn't only remain excellent resolution in frequency, but also avoid interference of pseudo difference signal for test results of wavelet transformation by morphological Arithmetic. For P and T wave, they are recognized by using slope of module maximum experiments are operating by database in Massachusetts Institute of Technology in Amecica(MIT-BIH), comparing with former results, nicer effect is obtained. Research expectation is given at the end of paper.