The precision of infrared temperature measurement instruments and the surface emissivity of the tested object have a great influence on the accuracy of infrared radiation measurement. In order to improve the calculation accuracy of surface emissivity, an infrared thermal imager is calibrated through a standard blackbody. Then, the temperature is measured with the calibrated infrared thermal imager, and the surface emissivity of the object to be tested is calculated. By using the calibration method of the infrared thermal imager based on neural network to calculate the target emissivity, the system error of the infrared thermal imager is effectively eliminated. The measuring equipment is simple and its measurement result is accurate. At the same time, the precise measurement of the temperature and emissivity has laid the basis for the development of infrared stealth materials.