心电波形识别,特别是P 波检测,要求较好地消除基线漂移。本文根据心电实时监护的要求,提出了分段三次函数法与分段抛物线法,在基线漂移校正效果与便于实时实现方面均取得了较好的结果。本文提出了自适应双阈值法作P 波的检测与定位,由34个正常人与心律失常病例实测结果,说明P 波正确检测率与P-R 间期基本符合率分别达到92.9%与87%。 ECG waveform recognition, especially P-wave detection, requires a good elimination of baseline wander. According to the requirements of ECG real-time monitoring, this paper proposes a piecewise cubic function method and a piecewise parabola method, and achieves good results in baseline drift correction and real-time implementation. This paper presents an adaptive double threshold method for P wave detection and localization. The results of 34 normal subjects and arrhythmia cases show that the correct coincidence rate of P wave and P-R interval is 92.9% and 87% respectively.