Abstract: For the needs of monitoring small high-speed moving object in space environment, a space target detection method based on infrared images was proposed to realize the detection, tracking and identification of non-cooperative space targets. According to the infrared characteristics of the target, the target detection progress adopted a threshold segmentation algorithm to segment the target. The center of mass of the target was extracted by the center of gravity method. The star background was eliminated by the use of the star pattern matching method. Target tracking and recognition process used trajectory correlation algorithm for multi-target batching and trajectory tracking. The target was ultimately identified using target classification algorithm based on the information of motion velocity and direction vector of the target. The experimental results show that with the proposed method, the tested space target can be detected within one second, meeting the requirements of good real-time detection, high detection probability, and low false alarm probability for space environment monitoring.