The Adaptive Threshold Adjustment Method Based on Event-Based Star Tracker

In order to solve the problem of difficult detection of star tracker caused by solar backlight in space environment, this paper proposes an adaptive threshold solution based on event camera, which includes a backlight Angle prediction model based on support vector regression (SVR) and a threshold optimization strategy based on receiver operating characteristic (ROC) curve. By analyzing the probe's historical data, the SVR model is able to make threshold adjustments for the sun's influence on the probe's field of view, while the ROC curve helps determine the optimal threshold to minimize noise effects and suppress extreme event rates, thereby improving detection accuracy in complex spatial lighting conditions. The results show that the proposed method can guarantee more than 96 % target detection capability and the event retention rate is less than 54 % in a 5 ms time window under indirect backlight environment. This adaptive method can suppress irrelevant optical noise while ensuring the detection accuracy, which proves its application potential in practical space detection missions.