The neutral winds in the mesosphere and lower thermosphere (MLT) region are measured by a newly installed meteor trail detection system (or meteor radar) at Chung-Li, Taiwan, for the period 10-25 November 2012, which includes the Leonid meteor shower period. In this study, we use the 3 m field-aligned plasma irregularities in the sporadic E (Es) region in combination with the International Geomagnetic Reference Field model to calibrate the system phase biases such that the true positions of the meteor trails can be correctly determined with interferometry technique. The horizontal wind velocities estimated from the radial velocities of the meteor trails and their locations by using a least squares method show that the diurnal tide dominates the variation of the MLT neutral wind with time over Chung-Li, which is in good agreement with the horizontal wind model (HWM07) prediction. However, harmonic analysis reveals that the amplitudes of the mean wind, diurnal, and semidiurnal tides of the radar-measured winds in height range 82-100 km are systematically larger than those of the model-predicted winds by up to a factor of 3. A comparison shows that the overall pattern of the height-local time distribution of the composite radar-measured meteor wind is, in general, consistent with that of the TIMED Doppler Interferometer-observed wind, which is dominated by a diurnal oscillation with downward phase progression at a rate of about 1.3 km/h. The occurrences of the Es layers retrieved from fluctuations of the amplitude and excess phase of the GPS signal received by the FORMOSAT-3/COSMIC satellites during the GPS radio occultation (RO) process are compared with the shear zones of the radar-measured meteor wind and HWM07 wind. The result shows that almost all of the RO-retrieved Es layers occur within the wind shear zones that favor the Es layer formation based on the wind shear theory, suggesting that the primary physical process responsible for the Es layer events retrieved from the scintillations of the GPS RO signal is very likely the plasma convergence effect of the neutral wind shear.