The distributed star sensor can be used as a monitoring platform for space targets. Multiple star sensors are needed to be associated instantly in the star map. The space target in the star map is a small target with no features (e.g., texture and color), which cannot be solved by the classical target association algorithm. Based on the correlation of polar geometry constraints, a space target synchronization association method under the distributed star sensor is proposed. The epipolar geometry constraint can describe the intrinsic projective relationship of multiviews, which has been widely used in image matching and stitching problems. It is applied in the target association field in this study. Based on the measured data, the association experiments of 300 star maps are carried out. The space target association probability can reach 90% and the feasibility of the algorithm is verified. The reasonable reference value of the associated threshold is proposed and various causes of the association error are analyzed. Monte Carlo is adopted to estimate the maximum probability of misjudgment. Experimental results show that the proposed method has an ideal association effect under the influence of jitter, measurement deviation and several interference points in the operation of the star sensor.