Detecting Asteroids in TESS Full-frame Images Using a Gradient-based Iterative Algorithm

Abstract The shift-stacking method is commonly used to detect asteroids in astronomical multiframe image data. However, the computation required for the traditional shift-stacking method is substantial. In this study, we observed a relation between velocity hypotheses and signal intensities in the stacked images. Based on this observation, we propose a gradient-based iterative algorithm to detect asteroids. The time complexity for detecting faint moving objects can be drastically reduced by avoiding calculations throughout the entire velocity space. A series of preprocessing steps were implemented to increase the effectiveness of the algorithm. We used the Transiting Exoplanet Survey Satellite full-frame image data consisting of 128 frames and covering a 12° × 12° area to demonstrate the performance of our pipeline. Within the data, 3143 independent detections were returned, of which 2266 are correlated to known asteroids. The completeness for asteroids brighter than 20 and 21 mag is 82.7% and 71.6%, respectively, and the detectability limit using our pipeline is estimated to be a visual magnitude of 21.8.