An Event Camera Tracking Based on MLS Surface Fitting Algorithm

Event cameras are different from CCD cameras, and event signal output only occurs when the light intensity is changed. In this paper, we propose an image processing method, first using the event-based visual flow algorithm to predispose the obtained data. Then use a method based on MLS surface fitting method and local plane fitting method to identify the images collected by the event camera. The method is tested in both uniform motion and high speed motion. It is concluded that this method can better filter noise and has high accuracy and robustness to edge identification and extraction.