Abstract: A dynamic simulation method for moving target recognition tracking and trajectory reconstruction in infrared thermal imaging video was proposed. Through the generation of virtual infrared images in the simulation environment and the basic model of imaging, a series of preprocessing was performed on the obtained images. A dynamic simulation platform based on Gazebo and OpenCV was built in the air-to-air scene. The smooth constraint algorithm was used to reconstruct the real-time dynamic trajectory of the tracking target. The error analysis model was proposed, and the performance of the trajectory reconstruction algorithm and the effectiveness of the simulation platform were analyzed. The experimental results show that this method has good accuracy and robustness for the trajectory reconstruction of infrared moving target in air-to-air scenario, and basically has no constraint on the motion model of the target. At the same time, the simulation platform has high operation efficiency and real-time performance. The real-time dynamic simulation above 60 fps can be realized by the ordinary household computer, which meets the performance test and training requirements of the trajectory reconstruction algorithm. The core algorithm can also be migrated to the airborne computing platform to realize the real-time trajectory reconstruction in the real scene. The proposed dynamic simulation method of moving target trajectory reconstruction in single-channel thermal imaging video is of great significance to the research of three-dimensional trajectory reconstruction and dynamic ranging and positioning of space targets.