Aerial video change detection boosted by visual servoing-based camera pose estimation, improving accuracy and reducing false alarms.
The article presents a new method for accurately estimating the position of a camera in aerial videos, which is crucial for detecting changes in images. The researchers used Visual Servoing to combine aerial videos with 3D models, allowing them to calculate the camera's position as it moves. By applying Newton's algorithm, they were able to determine the camera poses in the videos. The experiments showed that this approach is both robust and accurate, improving the reliability of detecting changes in aerial images.
