Curved Brick Building Facade Capture - Best Practices

Hi all,

I’m hoping some helpful community members can help me out with a difficult facade capture and reconstruction project I’m working on. We’re in the process of capturing the south (north is up) facade of the curved building shown below using an M210 RTK.

Our first pass at it captured the east 2/3 of the face with about 1600 photos using the Zenmuse X5S standard 15mm lens. The reconstruction has not gone as smoothly as I’d hoped, with very poor image calibration and matching, despite 50+ MTPs. I believe there are a couple of main reasons:

-Satellite lock in the area is sufficient for flight, but not resulting in pinpoint location accuracy for the processing engine

-The white brick is very uniform (a downsized sample photo is below). It does have plenty of stains/cracks/grime (the reason for the data capture!) but I’m not sure it’s sufficient for the Pix4Dmapper to pick up on. My plan here is to switch to a 45mm lens so that we can fly a bit further back and pick up additional satellites while also taking photos with more effective zoom so that the distinguishing features are easier for the processing engine to pick up on.

-Our flight pattern was in vertical runs, parallel to the building face, and achieved a high overlap. However, I’m thinking that having additional photos from further back may help the processing? Or perhaps a second run at another angle entirely. 

Any help or advice would be most appreciated. 

Hi Larry,

Indeed this project is very interesting, but at the same time, the building is tough to reconstruct. Could you send us the Quality Report of the processing you conducted? It would give us some reference to this project.

You can upload the QR to our OneDrive

Thank you in advance!

Thank you for the reply, Beata! I’ve started a support ticket and I’m working through some suggestions with David. I’ll follow up in this thread to share the results with the community.

Hi Larry!

It’s a great idea to share insights with other users!
Thank you in advance for your contribution. :slight_smile:

Best