Objective:To improve the detection of QRS complex in ECG signal with median filter,introduce extra criterions to determine the wave type and QRS boundaries. Methods:Based on the difference method,firstly noises such as 50 Hz noise and baseline drift were removed. After extracting amplitude and slope information of ECG signal,the ups and downs of the wave were obtained by adaptive thresholds. Results:Virtual detection system wes establish based on LabVIEW. The system was tested by MIT-BIH standard ECG database,and the correct rate reached 99.69%. Conclusion:Experiments show that the measurement is of high precision and lay the foundation for clinical ECG diagnosis.