Near-surface three-dimensional wind vectors are essential meteorological variables in atmospheric boundary layer research and environmental monitoring. Vehicle-mounted mobile wind measurement platforms enable observations of near-surface wind fields with high spatial resolutions over relatively large spatial domains. However, vehicle-induced flow distortion, platform motion, and attitude variations introduce substantial interference into vehicle-mounted wind measurements, resulting in systematic deviations of the measured wind vectors from the true ambient wind field. To obtain accurate three-dimensional wind measurements, this study proposes a vehicle-mounted instantaneous wind correction method based on a CFD-derived parameterization scheme. A set of parameterized CFD simulations is conducted within a unified computational domain under multiple wind speeds and directions to establish a mapping between the true inflow wind and the flow-distorted wind measured by vehicle-mounted anemometer. The correction framework further accounts for the effects of vehicle attitude variations and vehicle translation. Based on field comparison experiments, the corrected vehicle-mounted wind speed, wind direction, and vertical wind component show substantially improved agreements with ground-based reference observations under headwind conditions. Under tailwind conditions, the correction performance is reduced but remains superior to uncorrected measurements. Contribution analyses indicate that flow distortion is the dominant source of measurement error, whereas the contribution of attitude variations is generally smaller but can become more evident under low wind speed conditions or during more dynamic vehicle motions. The proposed method provides an efficient and extensible approach for vehicle-mounted three-dimensional wind measurements in complex near-surface wind environments.