Hi,
While scanning a building wall I found the accuracy of the perpendicular was about 7 cm. I used gcp targets on the walls and the z, up and down, and the left and right, which was NE - SW was good but the forward back was off. I only scanned one face. Would it have helped if I encompassed the building with the scan and more gcp targets on the walls.
Hi andreasprouse,
Are you able to share the project? I can send you a shared drive in a private message if able to share.
What make/model mobile device were you using for the capture, along with the version of PIX4Dcatch running on it?
When you say GCPs are these measured and marked points or are we referring more to manual tie points in this case?
I look forward to receiving more information.
Regards,
-Jon
Hi Jonathan,
I have a pile of things to do but I can send you something later this week. I am actually going to be at the site tomorrow collecting more data.
I used a Iphone 13 pro max and downloaded the latest catch.
The GCPs were targets we placed on the wall and measured x, y, and z with a robotic total station. I did place a couple manual tie points as well.
I’m thinking if I would have scanned another side of the buildings with targets it would help. Although I placed some high and some low they are all in kind of a straight line or a straight plane. If I can get this figured out it will be huge, being able to merge with our drone data.
On another note: Just flew a project near Orlando International Airport after getting an FAA waiver. Restricted to 25 meters. Just processed 849 images in Pixmatic, came out great.
Best Regards,
Andrea
Quality report attached along with an aerial
daytona aerial.pdf (2.8 MB)
DAYTONA WALL TARGETS-quality_report.pdf (32.1 KB)
Hi Andrea,
Thank you for the additional information.
I’m thinking if I would have scanned another side of the buildings with targets it would help. Although I placed some high and some low they are all in kind of a straight line or a straight plane. If I can get this figured out it will be huge, being able to merge with our drone data.
I believe you are saying that you are getting a flat plane with the marked GCPs on the wall. This would be expected and should be the case regardless of additional walls scanned. I can also read this as the GCPs are arranged in a straight line. If this is the case, they should be spread out and not in a straight line. I believe you have them arranged correctly, but mentioning just in case and for reference to future readers.
Another thing to consider when merging terrestrial and aerial datasets is to try and keep the two GSDs as close as possible. This will help when they are merged/processed together.
Happy mapping!
Jon