Newb question on uav maping

Hi Guys,

I am thinking of buying a reach RS set to use for ground control point collection.

I need these GCPs in order to accuratly map an area and to measure distance, surface and volumes.

I understand that averaging the base station point only give relative accuracy , meaning that the rover will shift by the base station shift.

My first question : Will relative accuracy be enough for GCPs collecion and mapping using pix4D ? Will I have cm-accurance distance, surface and volume measurements ?

My second question : What should I do to get absolute acuracy, I think it is important to know I live in Morocco ?

Thanks a lot for you help,

OK, where to start?
You turn up on site…
set up your reach base station.
If you have cellphone data and access to an NTRIP server or other source of corrections you can position your base in the reference frame of that system, otherwise it will be in WGS84 (ITRF2008). But no worries.
If you will be visiting the same area many times then make a permanent mark and set your base up over that.
get a rough fix (say 20min) and use the averaged position from this for your base station position.
(write those numbers down)
If you have a realtime link to your rover (3DR, LORA,…) then you can zomm off to each point with your base station running and make a good fix at each location and verify it there and then. If you dont have real time, again no worries. just log about 10 mins to half an hour of data at each of the GCPs.

Get out your laptops and follow the toolchain for post processing.
If you do get stuck send me a message and we can skype.

1 Like


Thank you for your detailed answer.
I do not have access to an NTRIP server.

I did read the post processing procedure using RTKLIB and RINEX logs -> Will this give absolute accuracy and cm level ?

Also, getting back to my first question : will relative accuracy be enough for mapping using pix4d ? will I be able to measure distance, areas,volumes…


Yes, relative accuracy will be enough. The accuracy of your volume, distance, etc. measurements will be the same either way.

Thanks Brent

So allow me to ask, what is absolute accuracy used for ?

Thank you

Well… there are lots of resources online that can explain that… better than I. Think of relative accuracy as a tape measure. Your are going to measure the size of a house. You measure length, width and height. All your measurements are relative to one corner. You start and measure 10m one direction, another 2m, etc, etc. These measurements are accurate and will give you an accurate measurement of your building, but they are only relative. They do not tell you where it is in the real world. With emlid reach, relative measurements are actually absolute measurements that do tell you APPROXIMATELY (3m accuracy or so) where you are in the real world, and then they make accurate measurements relative to the base station’s location. Absolute positioning provides a standard positioning system to people across the world. This way you can tell somebody else where something is in a standard system. So property lines, road construction, many different applications are suitable for this. Anything really, but it’s not always necessary. There’s my two bits.

1 Like

Great answer Brent, that is exactly what I was looking for.

So here is a resume (sorry I am beeing thorough before using this for my business):

  1. If using a UAV for drone mapping, and a reach system for GCP collection, we can achieve cm RELATIVE accuracy through the RTK capabilities of the system -> this is largely enough for doing measurement between objects in the same map.
  2. Absolute accuracy will only be needed in one of these cases :

If doing topographical work for cities, construction…, where the client needs to know exactly where the site is on the earth -> This is not my business
If mapping a large area isn’t possible in one flight: Then we have to do 2 or more maps, and to measure objects between different maps we need Absolute accuracy. But again, objects in the same maps will be measured with cm accuracy

-> Absolute accuracy can be obtained by using a Know point on the base.

Brent, is this resume correct for you ?

Again, thanks a lot for you help, this cleared up a lot of things, and conviced me to buy the reach RS and modules for mapping purpuses.

Best Regards,

In my opinion, it’s correct, except you CAN still map multiple flights. You just have to make sure you’re in the same co-ordinate system you used for the first flight, it doesn’t matter if it’s relative or absolute. Just leave the base station at the same place for measuring GCPs for both flights.
Achieving absolute accuracy isn’t that difficult, anyways. The very simplest would be to have a surveyor measure a point where you could leave your base station running. This would provide you with absolute accuracy anywhere with 10km. If you need to go farther, there are definitely ways to get that using the same Reach equipment. But if it’s not necessary, then it may not be worth it.

thanks a lot brent, I am going to post my summery addended with your remarks in the rtk part of the forum !

Great, One final thing.
If you are working without references station corrections in a known datum then you have to be clear what constellations you are using when you set up your base station. If the point you occupy will only be used by you then its easy, you can do what you like. If you turn all constellations on (can we do that or do we still have to choose between Beidou and Glonas?) then you will get a position that is in ITRF2000 reference frame to an accuracy determined by the length and quality of your observations.
The longer you can occupy a site for and log data, the better the absolute position. In post processing you can use the precise ephemeris files downloadable from ‘the web’ to improve the position even further.

My workflow for simple non repeat jobs with no corrections available is:
Turn up, put reference station on the roof of the car and turn on, ensuring logging is on. set the base mode to collect 15 mins of single fix data and average it then transmit corrections.
get everything else ready
turn on rover and use the transmitted corrections to get an RTK fix.
At the end of the day. Process the base station with the whole days data and compare to the 15min position. If the difference is greater than my required accuracy, reprocess the rover data with the new base position.

If I am likely to return to work in the same area. I will set the base up on a marked reference spot. Doesn’t have to be fancy, just permanent. Do the same process as above, but at the end of the day I WILL process all collected base data and fix the station properly. Then when I return I just plug those coords into the reach base station and my survey is repeatable. and locked into the abslute reference frame.

If I get real fussy I can combine multiple site visits together and reprocess the reference point, but this is getting a little to OCD.

From Navipedia (abridged)
The refined frame WGS84(G1150) was introduced in 2002, which agrees with ITRF2000 at the centimetre level.
the PZ-90.11 reference system was updated on all operational GLONASS satellites starting from 3:00 pm on December 31, 2013. From this time on, the satellites are broadcasting in the PZ-90.11[4]. This ECEF reference frame is an updated version of PZ-90, closest to the ITRF2000.
the initial realization of the GTRF called GTRF07v01 was in agreement with ITRF05 up to 0.9, 0.9 and 2.7 mm for North, East, and Up respectively.

1 Like

I am pretty sure I read somewhere in the docs, not to use beidou and glonass together…

1 Like

Thanks, thought that was the case. I skip Beidou.

Thanks for all these answers guys ! very helpfull and informative

This topic was automatically closed after 100 days. New replies are no longer allowed.