I have been struggling for a long time with the problem of creating a reliable algorithm that detects the movement of objects in a scene. I am constrained by the use of Yolo and a tracker (for example bytetrack).
My problem is to determine whether an object detected in the scene is stationary or moving, using data from the tracker such as ID, bbox position, etc.
As you can imagine, it should be able to cover a wide range of different situations and different motions, but my question, to begin with, is: Is there already an algorithm that does this kind of work?
Imagine that the tracker gives you for each object detected in the scene its ID, the position of its bbox, dimensions, etc., and that the algorithm, analyzing the data of each object, understands whether the object is moving or is stationary (over time).