Apply a data-driven feature tracking method, originally designed for event cameras, to aerial imagery to extract features and build feature tracks over an image sequence. These tracks will be used to estimate 3D camera poses using a Structure-from-Motion (SfM) algorithm. The project will also evaluate the quality of the recovered camera poses.
Develop a robust feature tracking pipeline for aerial imagery using event cameras. Build feature tracks from the aerial image sequence and integrate them with an SfM algorithm to estimate 3D camera poses. Evaluate the accuracy and reliability of the reconstructed camera poses.