In order to solve the problem that soil moisture sensor based on standingwave ratio (SWR) principle is affected by temperature in longterm operation, a temperature compensation model based on binary regression analysis was presented. Based on the least square principle, the parameters of the compensation model were determined and the sensor was compensated. The temperature drift characteristics of SWR soil moisture sensor were analyzed from two aspects of hardware circuit and the measuring principle of soil water content. The experiment using high and low temperature alternating humid heat test box set temperature in the range of 5℃ to 45℃, the test results showed that the absolute deviation of the sensor to measure soil moisture content volume was between -2.65% and 2.22%, the maximum relative error was 29.76%, and the maximum mean square error was 2.2119%. By fusing the SWR soil moisture sensor output value with the PT100 temperature sensor output value, the temperature compensation model was obtained by binary regression analysis based on the least squares optimization calculation, and the fitting determination coefficient was 0.998. The verification of the temperature compensation model depended on the sensor experimental data at different temperatures. The results showed that the absolute deviation distribution of the measurement results of SWR soil moisture sensor after temperature compensation was ranged from -0.26% to 0.69%, and the maximum relative error did not exceed 5.23%. The mean square error was decreased by an order of magnitude and the maximum was 0.157%. The temperature compensation model established can effectively reduce the influence of temperature on SWR soil moisture sensor and improve the accuracy and reliability of its measurement results.