Implementation of different triangulation algorithms

[aruna-ram]

Hello!

I will be using Caliscope to triangulate the position of a hawkmoth in flight from two camera views.

Since I am working with a two-view camera system, I was considering testing different triangulation algorithms—such as the optimal triangulation method for two views from Hartley and Sturm—to see if they affect the reprojection error compared to solving a DLT problem using SVD. (I believe this is what Caliscope does, but please correct me if I’m wrong!)

Would you be interested in integrating different triangulation algorithms or making it possible to do so using classes within Caliscope? If so, I could try integrating these methods into the Caliscope pipeline and perhaps submit a PR down the line. It may not be a priority now, but I wanted to share the idea! 😃

Thanks for reading!

[mprib]

I think this sounds like a great direction to take things. Gap filling, triangulation, and smoothing are all huge worlds unto themselves and it would be great to have more sophisticated approaches and options for users that integrate with the core camera calibration.