As an important part of space situational awareness, space-based surveillance of space debris is of great significance for understanding and protecting space resources. With the increasing number of spacecraft operating in the Earth's orbit, both the environment and the space security situation have gradually deteriorated. The accurate detection of small, dim targets in space is the technical basis of space-based debris monitoring; however, several difficulties exist in this field. Since space targets reflect low light energy, the number of pixels occupied by the target is small, and they do not reveal shape or texture characteristics. Space debris is therefore easily confused with background stars, leading to many false alarm targets, and making source detection, recognition, and tracking challenging. Accurate source positions for small and dim space targets are therefore important motivators in our project. However, the positioning accuracy of existing space-based monitoring platforms is limited by the position of the observation platform and the observable time period, due to the high-speed movements of the satellite platform and the observation target. The observation efficiency of existing space-based monitoring platforms is also low, which has certain limitations in practical applications. The high-precision measurement of target positions is thus inhibited when using existing space-based monitoring platforms. The aim of this research is to use an autonomous recognition algorithm based on a compound calibration star grid pattern for star identification. This study utilized a national advanced research project from the National High Technology Research and Development Program and used a star sensor for space debris monitoring, allowing dim and small space debris to be distinguished from stars. This research proposes a target detection technology path based on local sky reference star image comparison to detect weak and small targets in the background of star images. This technology enables the rapid, effective recognition and high-precision positioning of small, dim space targets.