Abstract: Abstract: Thin layer hot-air drying is one of the most important forms during drying, particularly with a wide range of applications. Among them, the moisture content is a significant quality index of drying products. Automatic weighing can be normally used to obtain the mass during drying, due to better universality, high accuracy, and simple architecture. However, temperature fluctuation, airflow disturbance, and mechanical vibration can posed the significant impacts on the weighing accuracy. This study aims to mitigate these impacts on weighing accuracy. A new online measuring device and the control system were developed for the moisture content during thin layer hot-air drying. Some features were automatically collected to timely view the data storage of material mass and moisture content in the drying process. The weighting measurement range was 0-2 000 g in the control system. The inverter, axial fan, and thermal air velocity sensor were also used to accurate control and detect of air velocity in the drying air velocity control system. The strain gauge weighing sensor AT8518 and weighing indicator XSB5 were then applied in the hardware of automatic weighing system. The Pt100 type temperature sensor was adopted to detect the temperature of the strain gauge weighing sensor AT8518. The weighing sensor was placed outside the drying chamber in the isolation from the drying environment, in order to increase the life span of the weighing sensor and the reliability of the drying device. As such, the direct interference of the drying environment was avoided to the weighing sensor. The frequency of the inverter was used to characterize the air velocity, in order to effectively avoid the measurement error of the thermal air velocity sensor in the variable temperature environment. The correlation coefficient of the fitting curve and the 95% confidence band indicated that the air velocity values were accurately converted from the current frequency of the inverter. Furthermore, the temperature fluctuation error of weighing sensor was calculated to establish the linear relationship between the normalized deviation of temperature change and weighing error. The measured mass was dynamically compensated by numerical correction as well. The constant load (987.16 g) calibration test showed that the maximum deviation between the corrected mass and the constant load mass was 0.368 g, and the mean average percentage error was reduced by 84.4%, the guide to the uncertainty in measurement was reduced by 72.2%, compared with the measured mass. Among the calibration values, the numerical range of 987.1-987.2 g was accounted for 77.8% of the total number of tests. A stop-air detection strategy was proposed to eliminate the airflow disturbance. Air velocity was calibrated with frequency of the inverter. The instant frequency of the inverter was recorded while weighing, and then the inverter was stopped for the complete clearance of the air velocity and mechanical vibration of the drying device. The values of mass were then continuously weighed to obtain the average value of the measured mass. At the same time, the numerical correction was conducted to obtain the corresponding moisture content. In the case of loading material of soybeans at 35, 45 and 55 °C, compared with the manual weighing, the absolute mass errors of automatic weighing were 0.337, 0.415, and 0.472 g, respectively, the absolute errors of moisture content were 0.141%, 0.304%, and 0.252%, respectively, and the root mean square error was less than 0.065, which fully met the requirements of online detection of material mass and moisture content in the drying process. The online moisture content measuring device for the thin layer hot-air drying was of great significance to evaluate and regulate the drying process for the agricultural materials. The finding can provide a strong reference for the high product quality and the low labor intensity.