Road crash laser scanners to save millions of pounds

I wonder why they use survey-style standing scanners. I'd put a standard reference marker on the ground and then fly a small quadrocopter UAV around the scene - for about $10-15,000 you can get pro ones designed to handle the weight of a pro camcorder and which are robust enough to operate in wind or rain.

The problem with the laser scanner on survey sticks is that you've got to do at least 3 scans and even then you're likely to have a lot of 'holes'. If there's a steep drop at the side of the road then too bad, you're not going to be able to take any readings from that side. Also, the height from which the readings are taken would be a limiting factor. It seems to me that a quality optical camera + laser scanner in the air could quickly deliver 360 degree coverage of the scene, followed by a zig-zag scan over the top. It wouldn't take any longer and could give you a much better model of an accident scene. The significant extra DSP required for all the image processing can be done on a computer in a police van or back at the station - it need not be realtime, because you can tell from looking at the stills/video whether you have good coverage or not.

I've thought this would be a great idea for a long time. Glad to see the technology is catching up and being implemented. My only concern would be that someone forgets a crucial angle or some level of detail is missed and then someone ends up blaming the new technology. Honestly you can miss information whether you're doing it the old way or with a scanner like this, but whenever you change processes, it's easy to blame the new technology.

wonder how this compares with a kinect.