[GH85Carrera]

Yea, sometimes programmers are expecting everyone to sit at the computer for endless hours watching the program run to hit the continue button.

On our mapping program that assembles the hundreds or thousands of images together into a single final image has three steps. Step one is just identify each image, and the position of the camera in XY&X plus the angle the camera was pointing, as in was the airplane going north or west or whatever, and the lens is ideally supposed to point down but the airplane yaws and rotates, in 3D. It is amazing the math needed to do that. Anyway the computer uses the camera GPS, and the IMU to make a best guess. It is usually within a few feet or up to 40 feet off. It also put it in the mapping coordinate system we told it we want to use.

We then go in and zoom in on a part of the image with a ground control point laid down by the surveyor and surveyed in to 1/10 of on inch accuracy. The target is often some 1 square foot black and white vinyl floor tiles. We tell our program that a certain point is the XYX the surveyor told us, and then we rerun step one. We have to then run a quality report to see the results. For whatever reason the program requires us to tell it manually, run the quality report. No way to tell the program, make the quality report every time. Often one or two targets need a little more tweaking to get them right where they are supposed to be. Once we get a eport that all is withing a very tight tolerance, we are happy and let it finish the next two steps.

In the end we get a "map" that the surveyors put their stamp on and send to the final client. It is a digital file anyone with mapping software from Google Earth to crazy expensive complex programs will open and place into the proper position on planet earth.