The purpose of this paper is to study the temperature measurement method of narrow band infrared camera. Firstly, starting from the field of image processing, an infrared thermal imager and a standard radiation source blackbody is used to collect blackbody infrared images at different temperatures at a certain ambient temperature. Then, the gray value of the blackbody image is calculated through mathematical modeling software and the correlation between the gray value and temperature is explored. Finally, the blackbody calibration curve is constructed based on the least squares method and interpolation fitting idea and the verification temperature is derived from the resulting calibration curve and the existing gray value. The validation results show that the temperature measurement accuracy has been improved, and the error is within 0.49 ℃.