Precise wind field data is essential for improving the accuracy of the numerical weather forecast, and tropospheric winds are not satisfied with the requirements as one of the key measurement objectives for improving weather forecasts. Although meteorological satellite-based imager derived winds by tracking the motion of characteristic targets in continuous cloud field images is an effective observation method that has improved numerical weather prediction forecasts on both regional and global scales, an error still exists in the ambiguity of the vector height assignment. Satellite-based infrared hyperspectral sounder has the capability of atmospheric vertical detection of temperature and humidity profiles, which can provide more accurate vector height assignment of wind field by analyzing atmospheric motion vectors among multiple vertical layers, improving the ambiguity of the vector height assignment. We proposed a method of tropospheric 3D winds measurement on cross-platform polar meteorological satellite-based infrared hyperspectral sounders of FY-3D/HIRAS and NOAA-20/CrIS, collocated vapor channel images through nadir overpass observations of both instruments, derived wind field by calculating the motion of dense optical flow field, combined ERA-Interim reanalysis data to verify the mean absolute deviation(MAE) and root mean square error(RMSE) of wind speed and the MAE of wind direction after quality control. The vertical wind fields of 200, 300, 400, 600, 650 and 1 000 hPa are calculated through observations of HIRAS and CrIS at 00:00, 06:00 and 12:00 UTC on February 20, 2019, the results show that, the trend of the variation of wind speed range is consistent with ERA-Interim reanalysis data, the wind speed range decreases as the height decreases, the upper layers are more sensitive to wind speeds above 20 m·s-1, while wind speeds measured near the surface are concentrated within 10 m·s-1. The MAE of wind speed is mostly less than 3 m·s-1 while the maximum value is less than 4 m·s-1, the RMSE of wind speed is mostly less than 3.5 m·s-1 while the maximum value is less than 4.5 m·s-1, the MAE of wind direction is mostly less than 30° while the maximum value is less than 40°. The wind field error mainly comes from the observation deviation of radiation data due to different instrument parameters, along with the positioning deviation of data image re-projection due to different spatial resolution.