I am new to PPK processing and would like to get a second opinion on my PPK results and workflow.
We flew a small area of 1sqkm to test our Powered Lift aircraft fitted with a modified Sony 5100 (Foxtech Map01 with 16mm Sony lens).
Before flying, we marked out 7 GCPs in the area with a Leica GS18 GPS unit.
We used RedCatch to process the aerial imagery and apply corrected geotags (in WGS84) from the Base and Rover logs.
We processed Step 1 of this imagery (image input WGS84 and output GDA94 / MGA zone 56) in Pix4D, ensuring camera optimization and matching was all good.
We imported GCPs (logged in GDA94 / MGA zone 56), changing all horizontal and vertical accuracies from 0.02 to 0.05 and marked 3x as 3D GCPs and 4x as Check Points, and Reoptimized.
Quality report showed excellent accuracy as shown under “Ground Control Points” section. Link to report here: Dropbox - File Deleted
To verify this, we ran Step 2 and 3 and used the Point Cloud to generate a surface model in different software. We re-marked GCPs on this surface model - including some extra independent check points not used in the pix4d processing and measured the points in the model. Again, RMSE showed just 19mm error in X, 13mm in Y and 30mm in Z - similar to the accuracies in the Quality Report (and within our target range of 50mm).
We are concerned these results are too good to be true!! What we would like some reassurance on is:
Does this workflow look correct for PPK processing?
Are there any other measures of accuracy we could perform to prove the level of accuracy?
Are we looking at the right metrics to judge the accuracy?
Perhaps this level of accuracy can be expected (and trusted) when using dual channel GNSS/survey-grade equipment?!
I have a WingtraOne PPK with the Sony RX1RII 42mp camera. I consistently get vertical RMSE accuracies of 0.05-0.10 feet (1.5-3.0 cm) without GCPs. I don’t even bother to check horizontal accuracy anymore, because vertical is always 1-2x GSD, so horizontal will be at least as good, or better. So all I do is collect topo check points. No need for targets.
So yes, if you have good photogrammetric and surveying workflows, you will get great results with PPK every time.
Hi Andrew I just purchased a Wingtra with the same camera. I am a land surveyor in Arkansas running ppk base using a trimble r 10 base. i get good ppk results after processing but i cannot for the life of me figure out why my checkpoints are not anywhere near the images. can you give me advice on the settings for input and output settings. my ellipsoid height in this area is about 93 feet below the orhto height(my msl ) in this area. thanks much. tony@and-survey.com
Andrew, I’m running an EBee but I want more accuracy. Have your run one? Are you a big fan of your wingtra? I’ve taken the second step to move to a base station but have wondered about PPK. Which do you think is better?
Tony, when you setup your base, what is the coordinate system that you are using?
I haven’t compared the Wingtra to an eBee, but I did compare it to a Delair UX-11, which is very similar to an eBee. I flew the same site at the same altitude with the same overlap using both aircraft. I processed both with Pix4DMapper using PPK data only, no GCPs. I compared the vertical accuracy of the DSM to 35 check points collected using RTK GPS. The vertical RMSe of the Wingtra was 0.086ft and the UX-11 was 0.153 feet.
So while the UX-11 was pretty good, the Wingtra was much better. I love my Wingtra.
What height did you fly at? I’m running at 170 right now and I am .1’ and under. This is with a base station plugged into the computer. With that being said, do you think PPK is more accurate? Is there a more accurate copter setup on the market? I’m looking to push my accuracies up and take away some of the noise from the Pix4D processing so I can have a cleaner point cloud. I’m also looking to process smaller projects, 1-7 house lots without taking landing damage for such a small flight. I’m just not exactly sure on how to do all of that yet. Quad with higher res camera… More GCP’s. Base vs PPK. Thanks for your help! I’m sure you’ve been doing this over and over again until you’ve found what is working best and I’m looking for a little short cut by asking for your knowledge.
400 ft agl. All of my research indicates that PPK is slightly more accurate than RTK. Also, PPK does not require a real-time connection, so the field setup is easier.
For a good multi-rotor, I would also look at Micro Drones.
Can you share with me what the issue is?
Although you are using PPK corrections you have 93 ft offset?
Can you tell us what are the steps you followed?
The issue is the vertical transformation you selected. Independent data collected using a GPS rover is usually referenced to a local Geoid to derive MSL elevations. P4Dmapper does not have the ability to automatically convert ellipsoid heights to MSL elevations using local Geoids (only global Geoids are supported). You have two options:
Calculate the geoid offset for your project location and enter in Pix4D. or
Convert the PPK data to local grid / elevation prior to import into Pix4D.
I am running the Wingtra with PPK. The corrections I use are my states C.O.R.S(continually operation reference system) adjustments. I believe my issue is the conversion between ellipsoid and ortho height. My problem there is the broadcast for our cors antenna does not have the difference between ellipsoid and ortho. I can run a base station for 2 hours for post processing, but in some instances I don’t have that option. I use the cors for my mobile LiDAR and it works like a champ so I know it is accurate. Is this a Pix4D conversion problem?
@JCash, as @Andrew_Milanes mentioned, in PIX4Dmapper you cannot select all the geoid models and this can cause the offset that you mention.
Besides the two workarounds that are mentioned (input the geoid offset, or convert the image coordinates in a third party software), you can also try out PIX4Dmatic that supports additional geoid model.
The possibility to specify the offset is available in PIX4Dmapper. This allows you to specify the offset between the ellipsoid and the desired geoid. The geoid that you work with is most probably not available in PIX4Dmapper so that is why we propose that workflow.
From the information I have at the moment, there is no plan to add additional geoid model directly in PIX4Dmapper so you need to do the workaround.
Since PIX4Dmatic has additional geoids available it might be suitable for your application.
These cookies are necessary for the website to function and cannot be switched off in our systems.
They are usually only set in response to actions made by you which amount to a request for services, such as setting your privacy preferences,
logging in, or filling in forms. These cookies do not store any personally identifiable information.
These cookies allow us to count visits and traffic sources so we can measure and improve the performance of our site.
They help us to know which pages are the most and least popular and see how visitors move around the site.
All information these cookies collect is aggregated and therefore anonymous.
If you do not allow these cookies we will not know when you have visited our site, and will not be able to monitor its performance.
These cookies may be set through our site by our advertising partner (Google).
They may be used by Google to build a profile of your interests and show you relevant adverts on other sites.
They do not directly store personal information but are based on uniquely identifying your browser and internet device.
If you do not allow these cookies, you will experience less targeted advertising.
