Correcting Orthomosaic or Point Cloud data using "corrected" base station

Hi All,

I mainly working with drone data both LiDAR and Photogrammetry at a forestry setting. In which both method used RS2 as a base station for PPK processing. We did not have a well distributed benchmark in our compartment but there is a continuously operating GNSS receiver that acquire data daily.

My proposed workflow would be:

  1. Static Processing the Base station in reference to the continuously operating GNSS receiver
  2. PPK Processing the sensor trajectory
  3. LiDAR/Photogrammetry processing

However, should the number one step cannot be done in order, can i use “average single” first to process the drone data and correct it after?

Anyone have try this? and how is the accuracy? is it the same with using the usual workflow?

Thanks.

There is no reason to use ‘average single’, unless you are flying with RTK. You stated that the plan was to use PPK, so all that needs to be done is store the raw gps data ( rinex and/or ubx) in both the base and rover. Then you can post process with corrected base coordinates whenever you get them.

You could do what you suggested and do a quick ‘average Single’ mission, fly mission with RTK using the ‘average single’ base coordinates, then post-process the mission whenever you get the base control point tied down good with static data.

Thanks for the reply!

What i mean is that sometimes the reference to tie the base station are not readily available and we cant delay the data processing to wait for the reference station data.

Are you familiar with correcting/offseting the processed drone data?

Absolutely. You can set your base to average single, fly the mission in RTK, then post-process the base position in whatever way you want (PPK, PPP) and then you can post-process your sensor data with the updated base position. In Emlid Studio, this would be by using the PPK worflow, setting the base position to manual and entering the post-processed position along with the raw data, then adding the sensor raw data.

Other post-processing software do the same thing with even more options.

Short answer is - Yes you should be able to.

I see what you mean. I do not have any experience with PPK on drones, but I have used PPK for multibeam hydrographic surveys, which is basically a lidar survey underwater - except that I use sound instead of light to measure range and there are no photographs. The similarity is that I also have an array of point cloud data that needs to be processed.

In my experience with PPK (on a boat), I did my initial survey using a commercial NTRIP service to get RTK positions while I was surveying. The governing body I was working for wanted two sets of data - one processed with the NTRIP and the other processed using PPK one of their benchmarks on-site. I first processed the NTRIP data applying all necessary parameters and ‘cleaned’ the data for bad soundings. After running the base station and rover’s (boat) GPS units though Emlid Studio (or some other 3rd party post-processing), I was able to get a POS file that I could load into my processed NTRIP data and replace the GPS device with the POS file.

IF you surveyed using RTK from the Single Average base, then you should be able a straight transformation of the data from initial base station coordinates used to the new static base coordinates.

Your workflow may be different, but there should be a way.

Thanks Gabriel! The idea is to give correction to the processed data without post-processing it all over again. Do you familiar with this workflow?

Hi Snillor999,

Thanks for the reply, it is really helping.

Have you ever tried the direct transformation on your data? Apply the same translation (x, y ,z) from the shifting of äverage" base station to the corrected position?

I want to incorporate it to my current workflow so i want to hear others experience on the matter…

In that case, just calculate the Δx, Δy, and Δz between your average single base and its post-processed position and apply this to your lidar point cloud through a global shift.

1 Like

I personally have not tried just a straight transformation. If I get some free time, then I’ll test it.

2 Likes

Thank you @snillor999 for the massive help! Looking forward for the result!!

1 Like

Hi there,

Just wanted to confirm what @snillor999 and @Gabriel_C said. You can update the data calculated relative to a single base position with its new coordinates, just like in this workflow.

1 Like