Real-time Change Detection between Drone Video and 3D Reference Model
Drones can be used to efficiently capture video over large areas. A post-processing mechanism called photogrammetry makes it possible to recreate 3D models from such videos.
In certain scenarios it is important to identify changes in the environment as quickly as possible – ideally as soon as the drone has captured the video.
A possible way to do this would be to compare the video stream in real-time against a previously generated 3D “reference model” of the same area.
The purpose of this Master’s Thesis is to study the feasibility of this approach, including performing the change analysis on board the drone itself.
This task is suitable for students interested in embedded systems and image recognition.