Say you have installed two cameras mounted on towers, separa…
Say you have installed two cameras mounted on towers, separated by some unknown distance, and aimed down at the same area of interest on the ground (images overlap each other). All interior orientation parameters for the cameras are known, but nothing else. Describe, in detail, a method using relative orientation and 3D coordinate transformation, to generate a dense, absolutely-oriented 3D map (point cloud) of the overlap area. Assume you have access to (can walk to) six small, widely distributed ground patches in the overlap area, but don’t have access to any other location in the overlap area. Almost all of the overlap area has texture and uniqueness suitable for automatic image matching. The only hardware you have access to is your office computer, a total station, and a base-rover RTK pair. Describe specific analytical photogrammetric processes: “Run the images through Pix4D” is not a valid answer for this question.