The invention discloses an electrocardiosignal ST section automatic judging method and device based on the artificial intelligent technology. For filtered human body electrocardiosignals, on the basisof wavelet filtering and a triangular area method, key character points of the electrocardiosignals are extracted, positioning S waves and T waves are included, the ST section initial point and slopecharacteristics are accurately recognized, then, a base line removal method based on mean value filtering is put forwards, extracted base line sequences are subtracted, novel electrocardiosignals areobtained, ST section and base lines are extracted for the electrocardiosignals, the heart rate and the slope of each section base line are calculated, the standard reference line is selected comprehensively, finally ST section abnormal changes are recognized according to the standard reference line, the qualitative and quantitative ST section judging result is obtained, and whether the base lineis lifted, lowered or normal is determined. The problems that due to base line shifting, base line errors are caused, standard base line selection is not accurate, and the abnormal heartbeat number difference is large are solved, the ST section abnormal change judging accuracy is indirectly improved, calculation is easy, and the method is easy to implement.