Abstract: It is difficult to accurately capture the target position due to the close strong interference and instantaneous occlusion interference between space targets and stars. Kalman filter is proposed to predict the target position and track the target, and the tracking gate is improved adaptive variable gate. At the same time, a new prediction and tracking method for strong interference and instantaneous occlusion targets is designed. When such targets are predicted, the predicted value of Kalman filter is used as the observation value to track the target, and the target prediction centroid is used to limit the correction of the gate centroid, so as to capture useful observation data. The experiment shows that the method can track strong interference and instantaneous occlusion targets robustly, improve the tracking accuracy of space targets, reduce data error rate, improve data quality, and provide more useful data for capturing orbital positions. This method runs fast and has certain applicability and scientific value.