Help needed

Compared to Google Maps or something else?

What problem are you referring to ?

I have requested access to your raw-files, then I’ll give it a try.

2 Likes

Hi. The blue dots are to the left of there actual placement when compared in pix4d (3rd picture) during processing.
I have been trying for over a year to get consistent, reliable data from them. When i post on here people say they have processed it and it was ok but no one will tell me what iam doing wrong.
since the last update they seem to have got worse, also i have started to have a battery issue that support are looking into on my reach RS+
I have shared the file with you

So, it is primarily your base position you are struggling with? Let’s try and solve this one step at a time.

2 Likes

What are the coordinates of the CORS station you are using?
Is there a website I can look up details on it?

2 Likes

Yeah. If you look at the first image it shows the location on google maps. i put the Long and latt into google. The second image is from the composite image of the site taken on the day. Allowing for the roads that have been put in it does not seem to be right. Its 11 pm here so iam off to bed but will turn on the PC first thing in the morning.

https://www.ordnancesurvey.co.uk/documents/resources/osnet-coordinates-file.txt found it here

2 Likes

Ok but what is it? Look like text from the .pos file.

It shows the precise and official position of the CORS station you are using.
What did you use for reference coordinate before?

Using EzSurv for post-processing, I get the following for your base:


You should have no problem reproducing this using RTKpost with default settings.

Looking at various maps, and comparing the drone images to these, I think the positions are the same. Draw a few lines on top of the old treelines/hedges and you should come to the same conclusion.
On top of that, then I don’t why the GNSS data should lie about your base position. The data seems sound, RMS is good.

In regards to the rover data, I can only get a fixed solution for Point 1. I can’t see a good reason for this, however, point 1 processes to a position that matched the GCP marking above.

Processing your rover file against your base file using RTKpost, I get a 99% fixed solution. From there, you can open the pos file in RTKplot and use Edit -> Timespan to manually input start/stop times (remember to account for the timediff between UTC and GPST)

1 Like

Morning pal. The problem is, if I’m honest is I don’t know where I’m going wrong.
I used the Blackpool airport CORS station (blap) from the CORS data from os.net. When i processed the logs i used emlid studio at the request from emlid support. This is the .pos file opened in RTKPlot and the location i got. (53.782600649 -2.772270354 in case the image isn’t clear) is not the same as yours.


When i process the rover logs using the base co-ordinates I get this. Which gives a 99.7% fix.

According to the CSV file below Point one i started the recording at 2021-07-01 10:54:56 UTC and stopped at 2021-07-01 11:05:06 UTC

Allowing for the difference between UTC and GPS time I used time span interval and started at 10:55:14 and stopped at 11:05:24 and got this. Yes the RMS is great but its not the same as the actual location of the GCP when put into Pix4d
.
This is the location for Point 1 when i put the co-ordinates into google maps it shows as here as the red icon. i know google is not that accurate but even if you take that into account it no where near the actual GCP location.

How ever when you look for the GCP on the composite image it shows up in the red circle. I have over laid the google image on to it you can see there not in the same location

The image below was taken from Pix4d while processing. the blue dot is where pix thinks the gcp is according to the .CSV file, however it cant finish processing as the GCP is actually where the yellow arrow is. All of the points in the CSV file have the same error, they are all over to the left when looking at the image and all by roughly the same distance.

I don’t know if it something I’m doing wrong or if the units are faulty but I’m at the point where I am going to cut my losses and bin them so I can get something else that wont cost me clients or reputation. in total I have spent around £1500 and I am not getting reliable accurate data and no one can tell me why. Its so frustrating.

What reference coordinate did you use for this processing?

I don’t have any reason to believe that the units are faulty, as I was able to get the correct positions.
I think your CORS reference coordinate is incorrect (compare to the link to ordnance survey I sent earlier), and thus all resulting coordinates are also incorrect, as all processing are derived from this coordinate.

Hi pal the reference position came from the .pos opened as text and its not far from yours but its here.

Yeah, that is practically the same coordinate for the CORS station.
What coordinate did you use for your own base, when you process up against your rover?

I used the coordinates from the above image
53.782600649 -2.772270354 79.9234
These are further away from yours but still not that great

That seems reasonable the same. Not too much of difference.
So, we have established that none of the base coordinates seem to be faulty.

So it must be at the step of calculating the rover kinematic solution. Can you try and process with RTKpost instead?

I will do. but to avoid any further confusion can you show me the setting you used. That way we know its not some i do during that and we both have the same settings.

Sure thing:
RTKPostSettings2 RTKPostSettings1

just processing so ill let you know.

1 Like

Well first impressions of the new coordinates are that they seem a lot better. They all seem to be in the right area. There were only a few differences in the setting but they seem to have made a difference.
The Max pos VAR for AR i had set to 0.3000
Frequencies i had set to L1
Filter type was set the forward.
Iono/ tropo corrections and Sat Ephemeris were all different.
A friend is processing it in Pix4d so ill get the results this afternoon

1 Like

Sounds good ! My settings shouldn’t be too far off defaults, as I remember them!

Well it made all the difference, that a some degree of user error. The last CSV File processed a 22mm in Pix4d Thank you for your help. I will try it out again in a different location and with more GCP’s THANK YOU FOR YOUR HELP.

2 Likes