Activity Feed › Discussion Forums › Photogrammetry, LiDAR & UAS › Look Ma, I’m a Photogrammetrist
3 times? Not familiar with that. Do you have a link?
7.8 Accuracy Requirements for Ground
Control Used for Aerial Triangulation
Ground control points used for aerial triangulation should have higher
accuracy than the expected accuracy of derived products according to
the following two categories:
? Accuracy of ground control designed for planimetric data (orthoimagery and/or digital planimetric map)production only:
RMSEx or RMSEy = 1/4 * RMSEx(Map) or RMSEy(Map), RMSEz = 1/2 * RMSEx(Map) or RMSEy(Map)
? Accuracy of ground control designed for elevation data, or planimetric data and elevation data production:
RMSEx, RMSEy or RMSEz= 1/4 * RMSEx(Map), RMSEy(Map) or RMSEz(DEM)https://cdn.ymaws.com/www.nsps.us.com/resource/resmgr/Model_Standards/sectionf.pdf
6. MINIMUM HORIZONTAL ACCURACY
The horizontal accuracy is based upon the American Society of Photogrammetry and
Remote Sensing (ASPRS) Standard for Class 2 and reported in agreement with the
National Standard for Spatial Data Accuracy. The NSSDA Horizontal Positional
Accuracy Statistic at the 95% confidence level is determined by multiplying the Root
Mean Square Error (RMSE) of the data set by 1.7308.
Acceptable
Base Mapping Scale of LIS/GIS Positional Accuracy Statistic of Survey Data
1?= 20 ft. 0.7 feet
1?= 50 ft. 1.7 feet
1?= 100 ft. 3.5 feet
1?= 200 ft. 6.9 feet
1?= 400 ft. 13.8 feet
1?= 500 ft 17.3 feet
1?= 1000 ft. 34.6 feet
1?= 2000 ft. 69.2 feet
3
7. MINIMUM VERTICAL ACCURACY
The vertical accuracy is based upon the ASPRS Standard for Class 1 and reported in
agreement with the National Standard for Spatial Data Accuracy. The NSSDA Vertical
Positional Accuracy Statistic at the 95% confidence level is determined by multiplying
the Root Mean Square Error (RMSE) of the data set by 1.9600.
Acceptable
Base Mapping Contour Interval Positional Accuracy Statistic of Survey Data
1 foot 0.7 feet
2 feet 1.3 feet
5 feet 3.2 feet
10 feet 6.5 feet
15 feet 9.7 feet-All thoughts my own, except my typos and when I am wrong.@drew-r – Up until approximately 8 years ago, the California Board used another state (printing) agency to print the wall certificates. This changed when the Board began to hear from licensees that certificates were printed incorrectly including out of alignment. Some of the returned certificates should had never gone out like they did. (engineers will take the time to slap a scale on these things to check!) There is no valid excuse for misspelling a discipline type. The Board assumed control and now prints all wall certificates in-house.
@ric-moore I got my license in 1984, which I believe is the first (and last?) year the CA Board decided not to print the licensee’s name in a nice black Leroy-style font, instead hiring a calligrapher to write the name in an artsy-fartsy semi-script. I was quite disappointed that my license didn’t look like those of my mentors, instead it looked like an award from a youth summer camp. And the calligraphy has faded badly over the years; still legible, but pretty faint.
They literally say “three times”? The above is directly from (and linked to) the NSPS publication.
I have no idea what “three times” accuracy even means. Can you explain that? Does that mean 3 times the order of accuracy? Does that mean .1′ instead of 0.3′? Because that is a heck of a lot more accurate than 3 times.
Seriously, I am curious where that is written, I am always open to learning.
Link?
Based on what I see in the link I have, the accuracy for 1 foot contours is well and truly exceeded by (3?) many times the accuracy required (0.7′) if you used RTK.
-All thoughts my own, except my typos and when I am wrong.Took me a while to find it, since I just now noticed NSPS updated recently 2021. My bad. Appears this info is out of date and was only a draft.
ASPRS POSITIONAL ACCURACY STANDARDS FOR DIGITAL GEOSPATIAL DATA ?? ASPRS
Draft_ASPRS_Accuracy_Standards_for_Digital_Geospatial_Data_PE&RS.pdf
We will see up to a tenth of variation in the same session, but we will see easily two to three tenths difference when returning to the same points only a few hours later. This is with a network, not single-base solution, and pretty much anywhere within the coverage area.
Even around here with a well-run RTN and plenty of stations around running full GNSS, this is just how it is. If I want repeatable vertical with respect to the network, we’re doing multiple 1- to 2-minute observations spaced out over a few hours.
Then again, for local project control that only need to be tight within itself, a single visit will suffice to bring us within tolerance, and then we can tighten things up with TS and level work. Most surveyors claim that sub-tenth accuracy because they only observe with GNSS once, at the same time of day, and never again.
“…people will come to love their oppression, to adore the technologies that undo their capacities to think.” -Neil PostmanThere is no such thing as a VRS base “five miles away” as far as I know. That is why I was asking…I was trying to fill in my ignorance. The software creates a Virtual Reference Station at the location that you start your session. So, as far as I know there is no reason to ever have a base very far away at all. Depends on what you mean by “base”, I suppose.
-All thoughts my own, except my typos and when I am wrong.@leegreen – is it not easier to just use the aerial GCP’s you set and log their locations in with either GNSS or a total station if the site is smaller to mid size and just do some random spot checks for the topo on the asphalt, dirt and grassed areas? seems like survey grade equipment would work better anyway than DJI’s stuff plus the drone is constantly moving & drifting I would think. I don’t have any personal experience so that’s why I ask
Correct. The VRS creates a virtual base at your initialization point. However. if you reset your VRS, what is the repeatability of your position when the Network CORS is more than 5 miles away?
When you return to a site the next time, can you use the same VRS initial point?
Yes, using the correct amount of GCP’s will be just as accurate as PPK or RTK drone. The DJI P4 Pro and DJI P4RTK have the same camera, so yes same results once registered correctly. With PPK and RTK you still want several GCP’s or checkpoints for ground-truthing.
@leegreen – thank you sir for the clarification! curious how your results for topographic work come out? I would think that the density of the point cloud would help tremendously as compared to conventional methods
Where there is very little texture in the surface over a large area then the PPK/RTK will reduce the risk of a photogram mismatch causing a shift on a number of overlaps, so you don’t need to intensify the density of GCP in those areas. The texture problem is exacerbated in poor light (when you shouldn’t be flying – but occasionally needs must!).
Log in to reply.