The performance evaluation of the airborne infrared imaging system is very important for the task performance of aircraft. modulation transfer function (MTF) is a typical index to measure the performance of the infrared imaging system. At present, MTF measurement depends on specific image features and manual operation. Hence, it is difficult to achieve real-time and batch evaluation. In this article, an automatic MTF measurement method based on non-specific images is proposed. The method can automatically obtain MTF values from actual scene images without any specific features, which is realized through image edge detection, target region extraction, and the improved slanted edge method. Then, an equivalent physical model of the infrared imaging system is established to achieve MTF curve fitting and filter out random errors. Finally, a certain type of UAV infrared imaging system is taken as the verification object, and the absolute error between the MTF measured in the laboratory and the MTF calculated based on non-specific actual scene images is less than 0. 06, which proves the feasibility and effectiveness of the proposed method. This automatic measurement method not only can evaluate the MTF accurately in real time, but also provide a new performance testing idea for infrared imaging system which is inconvenient to conduct ground experiments.