Improve results with LiDAR or Selected Features?

[avclubvids]

Describe the problem I have a calibration capture from a camera array that used Agisoft tags for their calibration. I want to try to get better results in Meshroom but there is no support for Agisoft's tag format as far as I know. I have a LiDAR scan of this capture, so I have millions of accurate 3D points in a point cloud. Is there a way for me to do any of the following:

• detect Agisoft tags
• manually select 2D features in individual images or manually associate SIFT results between images
• assign detected 2D features with known 3D locations in the LiDAR cloud

I would love to be able to use the LiDAR data to improve accuracy but I currently do not see how that could work.

[natowi]

Point 1 would require some implementation affords. If someone is interested, here is an example on how to add a new marker type tom Alicevision: https://github.com/alicevision/AliceVision/pull/950 Point 2 and 3 are on the to do list but not yet implemented.

[avclubvids]

@natowi - thank you for confirming that it was not just user inexperience on my part :)

I do not think that there is a way to implement Agisoft Tag detection without violating some EULA or patent of theirs so I don't think that's a path forward even if I had the coding chops to do it myself.

As for options 2 and 3, are there issues assigned to these tasks or is there some other place where I can follow along and participate in helping to refine and finalize these functionalities?

[natowi]

The patent for the Circular Tag used by Agisoft recently expired: patent DE19733466A1. https://patents.google.com/patent/DE19733466A1/ https://register.dpma.de/DPMAregister/pat/register?AKZ=197334660 Here is some code: https://github.com/natowi/14-Bit-Circular-Coded-Target and here https://github.com/poxiao2/CCTDecode if someone is interested.

If you have coding experience you are welcome to help. Best write an Email to team@alicevision.org to discuss details.