Drone Mapping with Leica RTK GPS system | Help

Joined
Jan 7, 2021
Messages
1
Reaction score
0
Age
28
I recently got into drone mapping and invested a great deal in all the equipment needed. The gear I'm using is the Phantom 4 Pro V2 with the Leica GS18T rover with a base station and have some landing pad GCPs.

In short, everything went smoothly until I imported the GPS points I captured with the Leica unit into Pix4D software and it was just miles off from the testing site.



This is the workflow I followed during my test:

1) Plan flight with DroneDeploy

2) Arrived to field. Turned on base station (Leica GS16) and rover (Leica GS18T). Made sure to get an accurate reading down to centimeter-accuracy. The base had a few options for the setup which is: set "over a known point," or "over any point." I chose over any point. Coordinate system was set to my local stateplane coordinate system NAD83 MA11. (I'm based in Guam)

3) I laid out 5 GCP pads evenly across the field and went to each one with my Leica rover to collect the center point.

4) After points were collected, GPS units were packed away and I proceeded to the drone flight plan via DroneDeploy.

5) Go back to office, start a new project in Pix4D, upload the photos, set the coordinate to the state plane, process step 1 and import the GCP coordinates via GCP/MTP manager.

6) This is where it goes all wrong. It immediately gives me a prompt saying one or more GCP points are far away from your site. When I click "okay" to just proceed anyways, the GCP points are miles away from the the testing site.



I'm pulling my hair not knowing what went wrong and been trying to figure this thing out for two weeks now. What I'm assuming I'm not doing correctly is the workflow with the GPS unit?

I've been reading a lot into this and I hear surveyors typically position their base station directly over a "known point" or a "survey monument" when they're capturing coordinates with their rover.



Some burning questions I'm itching to get answers for:

1) Am I supposed to place my base station directly over a "known point" or "survey monument" for every job if I want to get accurate measurements?

2) Or am I doing something else wrong like the coordinate system settings?

3) What is typically the workflow for drone mapping with an RTK GPS system? (RTK system meaning base + rover. Unfortunately accessing the NTRIP network is not available in Guam apparently)




I would really appreciate any insight as to what I can do to correct this issue.
 
1) Am I supposed to place my base station directly over a "known point" or "survey monument" for every job if I want to get accurate measurements?

It's always a good idea to use a known point (if the known position comes from a reliable source). It takes away the possibility of a "bad fix" with the GPS. Any time you set your base on an unknown position, the GPS unit must be given enough time to fix to enough satellites to obtain a fair position. Without any post-processing of the data from the base station and a proper fix, you're likely going to obtain coordinates that are within 10 meters of a true position. Download the data file from the base station and process it through the NGS OPUS online system. OPUS will process the data file and return by email a refined position in the sub-meter category or better if the base observations were long enough (15 minutes to 2 hours for a rapid static (floating) solution or 4 hours to 48 hours for a static (fixed) solution). If you're finding yourself off by miles, then you possibly have a problem with the coordinate system you're using either in the field or the PIX4D processing or you have inconsistent units (ft/m).

2) Or am I doing something else wrong like the coordinate system settings?


GPS isn't a magic catch-all simple flip-a-switch solution in a box. Without the proper knowledge of geodetic coordinate systems and training on the mathematical principles involved, it's easy to make a mistake. Professional land surveyors train for years to obtain a working understanding of geodesy involved in the GPS system. Look closely at the data file to confirm the coordinate values to see if they are reasonable for the coordinate system you've chosen to use. It also helps if you pull the lat/long position from Google Earth and convert it to the proper SPC to compare with your base results to see if you're in the ballpark. The NGS website has tools available to do the conversion.

3) What is typically the workflow for drone mapping with an RTK GPS system? (RTK system meaning base + rover. Unfortunately accessing the NTRIP network is not available in Guam apparently)


When using an autonomous (unknown) base position, make sure to let it "cook" long enough for the position to settle down before you "lock-in" the base position. When you first turn on the base, watch the position it's returning closely on your data collector to make certain that it's settled within acceptable norms (slowly wandering within a 10-meter circle without radical jumps) before you start your rover. Another possibility is to set up your own NTRIP system by fixing your own base station at home or work. You'd be fine with sub-centimeter positions within 100 kilometers of your base. It just adds another level of complexity to your GPS system.
 
I recently got into drone mapping and invested a great deal in all the equipment needed. The gear I'm using is the Phantom 4 Pro V2 with the Leica GS18T rover with a base station and have some landing pad GCPs.

In short, everything went smoothly until I imported the GPS points I captured with the Leica unit into Pix4D software and it was just miles off from the testing site.



This is the workflow I followed during my test:

1) Plan flight with DroneDeploy

2) Arrived to field. Turned on base station (Leica GS16) and rover (Leica GS18T). Made sure to get an accurate reading down to centimeter-accuracy. The base had a few options for the setup which is: set "over a known point," or "over any point." I chose over any point. Coordinate system was set to my local stateplane coordinate system NAD83 MA11. (I'm based in Guam)

3) I laid out 5 GCP pads evenly across the field and went to each one with my Leica rover to collect the center point.

4) After points were collected, GPS units were packed away and I proceeded to the drone flight plan via DroneDeploy.

5) Go back to office, start a new project in Pix4D, upload the photos, set the coordinate to the state plane, process step 1 and import the GCP coordinates via GCP/MTP manager.

6) This is where it goes all wrong. It immediately gives me a prompt saying one or more GCP points are far away from your site. When I click "okay" to just proceed anyways, the GCP points are miles away from the the testing site.



I'm pulling my hair not knowing what went wrong and been trying to figure this thing out for two weeks now. What I'm assuming I'm not doing correctly is the workflow with the GPS unit?

I've been reading a lot into this and I hear surveyors typically position their base station directly over a "known point" or a "survey monument" when they're capturing coordinates with their rover.



Some burning questions I'm itching to get answers for:

1) Am I supposed to place my base station directly over a "known point" or "survey monument" for every job if I want to get accurate measurements?

2) Or am I doing something else wrong like the coordinate system settings?

3) What is typically the workflow for drone mapping with an RTK GPS system? (RTK system meaning base + rover. Unfortunately accessing the NTRIP network is not available in Guam apparently)




I would really appreciate any insight as to what I can do to correct this issue.
It does seem like the coordinate readings are wrong. is it possible that teh Leica got set to lat/long or UTM even though you tried to set it to State Plane? I'm sure you'll have checked that, but it would be the explanation that best fits your observations. OR, I thought the whole benefit of "Landing Pad" control points was that they broadcast their own coordinates. Is it possible that they are still/only in WGS84 lat/long and somehow overriding the coord. you collected with thje Leica? (NOTE: I'm not a surveying professional and have no experience with such fine equipment as you have; I just teach GIS and see these issues with students 'management' of spatial references frequently)
 
  • Like
Reactions: JBStahl
It does seem like the coordinate readings are wrong. is it possible that teh Leica got set to lat/long or UTM even though you tried to set it to State Plane? I'm sure you'll have checked that, but it would be the explanation that best fits your observations. OR, I thought the whole benefit of "Landing Pad" control points was that they broadcast their own coordinates. Is it possible that they are still/only in WGS84 lat/long and somehow overriding the coord. you collected with thje Leica? (NOTE: I'm not a surveying professional and have no experience with such fine equipment as you have; I just teach GIS and see these issues with students 'management' of spatial references frequently)
Good points. It could also be that the lat/long coordinates are set to decimal degrees in one system and degrees/minutes/seconds in the other. Problems caused by feet to meter conversions usft vs international ft may also cause positioning errors (but not usually measured in miles). I'd also be very cautious when using UTM for anything related to real ground precision as the displacement from geodetic to reality can get way off in orientation and position.
 
Good points. It could also be that the lat/long coordinates are set to decimal degrees in one system and degrees/minutes/seconds in the other. Problems caused by feet to meter conversions usft vs international ft may also cause positioning errors (but not usually measured in miles). I'd also be very cautious when using UTM for anything related to real ground precision as the displacement from geodetic to reality can get way off in orientation and position.
Yes. dec. deg. vs. dddmmmsss seems more likely for 'miles off' rather than 000s of miles off...
 
If you are concerned then survey in four boundary stakes and see how they match up with the rtk . You should be be with in the register surveyor standards
 

Members online

Forum statistics

Threads
143,066
Messages
1,467,358
Members
104,936
Latest member
hirehackers