In my disscussiod with colleagues who specialize in aerial photogrammetry.?ÿ They say the more panels the better.?ÿ However one did mention the point of deminishing return.?ÿ When too many points and?ÿ
observations of more than 4 or 5 hours might result in less accurate results.
Introducing a lot of extra GCP's has only brought the accuracy down in my experience.
As others have implied, I think the phrase "brought the accuracy down" means "showed that the error estimates of a less well-controlled survey are unrealistic."
Introducing a lot of extra GCP's has only brought the accuracy down in my experience.
As others have implied, I think the phrase "brought the accuracy down" means "showed that the error estimates of a less well-controlled survey are unrealistic."
I've spent nearly 25 years rectifying imagery with GCPs.?ÿ Most of that has been for geologic exploration using satellite imagery.?ÿ There are instances where artifacts in the imagery will create problems with the rectification and having too many GCPs can create "tears" in the image.?ÿ More often than not this occurs with poor geometric distribution of the GCPs.
The other thing I'd like to mention is that the "extra" GCPs do not have to be used as control points to create the topographic model.?ÿ They can be reserved to evaluate the accuracy of the bare earth model created by the software.?ÿ Just like redundancy in least squares is a good thing, having a sufficient number of elevation residuals to evaluate the accuracy of the model is a good thing.
I could see more targets making for higher residuals, but not lower accuracy. Unless the targets themselves weren't accurate.
I know nothing about UAV topo mapping but have considerable experience in traditional (pre-GPS) photogrammetric mapping and would like to conjecture why UAV mapping requires so many GCPs.?ÿ
It's basically that UAVs fly low (400'+-) and airplanes fly high (2,500'+-) so a UAV job generates hundreds of "photos" compared with a few dozen photos using traditional photogrammetry for a site of size "X".?ÿ ?ÿThe required number of GCPs is basically a function of?ÿ the number of photos that have to be stitched together so to speak, so a UAV job would require an order of magnitude or more GCPs.?ÿ Also, a photogrammetric camera is an exquisite optical instrument, a UAV camera, not so much, so more GCPs would be required on a UAV job to better control camera distortion.?ÿ
I was puzzled by an above poster's statement that "I typically fly 150 acres with 6 GCP's and typically get a verified accuracy of 0.1' " but indeed, a quick surf revealed many UAV mapping firms quote those specifications.?ÿ Is the key that the UAV is using Airborne GPS/inertial to determine the location and platform orientation, etc., of the camera??ÿ Similarly, airborne GPS/inertial lowers the number of GCPs required in traditional photogrammetry, but the equipment involved is apples to oranges;?ÿ can a UAV carry a GPS that compares with the survey grade receiver(s) and ground base station(s) an airplane utilizes??ÿ Or is it simply that the UAV is low and slow compared to an airplane and doesn't need as accurate location/orientation data?
I'm really puzzled by the original poster's statement of 70 GCPs for 60 acres, and a replier's statement of 6 GCPs for 150 acres.?ÿ Somebody's going out of business if both are true.
If your panel placement and data is good you cannot reduce overall dsm or photo quality. What it can do is give more realistic statistics and maybe a better product. I would encourage users to get the latest Wolfe book for getting a handle on the numbers.
Regarding the 'insufficient description '...
I end those with a declaratory statement of, 'this description is intended to include all of lot x, together with adjacent vacated r-o-w and no other property '..?ÿ
For some reason I dont get notifications to threads I respond to. I'll be back tonight/monday with more answers as Im about to run into a meeting but its basically a combination of a couple things.
Its a little bit of what Gene Kooper said:
"There are instances where artifacts in the imagery will create problems with the rectification and having too many GCPs can create "tears" in the image.?ÿ More often than not this occurs with poor geometric distribution of the GCPs."
And aso what Jim Frame has said:
"As others have implied, I think the phrase "brought the accuracy down" means "showed that the error estimates of a less well-controlled survey are unrealistic."
?ÿ
90% of the flights I have done, I've gone back in the field to check the elevations from the UAV topo. its always within 0.1' with my 6 GCP's.
?ÿ
Also to answer another above poster - yes my UAV has onboard GPS to locate the photos. It isnt RTK - although you can get them with RTK - but accuracy similar to your phones GPS. Which is enough to get locations for the photos so you can run the software.
The City can't convey property owned by others by Ordinance.
Tell that to them!?ÿ It's like talking to a wall. I should of enrolled in a Dale Carnegie class. :)?ÿ Jp
As far as the density of the targets, I was told that they are to prove the accuracy of the topo. There were some positional accuracy limits?ÿon the last one we did for them and it was tight.
I didn't pay much attention to all of that since we could reach it on our end without much effort, most of the check shots on the targets were?ÿabout .02' vertically.
The site was small enough and there were so many targets I kinda figured we could have run it with the 4 wheeler as fast as laying out, locating twice and picking up the targets.
But then there wouldn't be a pretty picture.
And easy money, which is the best part ??ÿ
When powers that be won't listen:
1. Have a drone survey to do: not a big deal, the drone people hired us to lay out around?ÿ70 targets over a site, this is a campus and I figure it will go smooth. Had it set up for a couple of weeks ago on a Sunday while there is nothing going on. The powers that be for the campus decided it was too early and want to wait a couple of months because there is something they want seen in the photos. Meanwhile the trees are leafing out and it's going to for sure degrade the survey in the areas that were open when we wanted to do it......
?ÿ
Who are the drone people? Are they a separate company that creates their own map and only hires the surveyor for the GCPs? Or are they another survey outfit that subbed out the control to you? Just curious...
When powers that be won't listen:
1. Have a drone survey to do: not a big deal, the drone people hired us to lay out around?ÿ70 targets over a site, this is a campus and I figure it will go smooth. Had it set up for a couple of weeks ago on a Sunday while there is nothing going on. The powers that be for the campus decided it was too early and want to wait a couple of months because there is something they want seen in the photos. Meanwhile the trees are leafing out and it's going to for sure degrade the survey in the areas that were open when we wanted to do it......
?ÿ
Who are the drone people? Are they a separate company that creates their own map and only hires the surveyor for the GCPs? Or are they another survey outfit that subbed out the control to you? Just curious...
They are an engineering company, one of the engineers does the flights, they have?ÿa PELS on staff who usually will do control, but he has been half retired the past year or so and isn't always available, so we fill in sometimes.
Hey easy money is easy money, but if the site is that small then they definitely are setting way way too many targets.
