Using free & faster AI based COLMAP instead of CaptureReality or Meshroom for Photogrammetry

Development Asset Creation
, , , ,
1

Hi there,

I am working to use the absolutely revolutionary software COLMAP that uses AI to Turn 2D images into 3D scenes. It seems to be about 1600 times faster than Meshroom and about 1000 times faster than CaptureReality and it gives much higher quality results and is free.

https://colmap.github.io/

New Scientist article here: Artificial intelligence: 2D images transformed into explorable 3D worlds | New Scientist

With a great demo video:

I am having a problem based on my very low levels of development education and capacity and was wondering if anyone could help please? The software is going through a crash that they anticipate and describe solutions for in the documentation regarding the memory of the GPU being used up Frequently Asked Questions — COLMAP 3.8 documentation and it requres sending commands to the program from powershell but I can’t for the life of me work out how to do this. When I open the Program with the Powershell it wont let me enter in any more commands, it just stops working. Please can someone help? I think this is the future of volumetric capture. Many thanks,

2

I think you fundamentally misunderstand what COLMAP is. COLMAP is a general-purpose SfM solution. It is not AI-based or AI-assisted. In fact it is much slower than CR, Meshroom, or Metashape at determining camera positions. I suggest you do some further research.