Dealing with matrices: Eigen vs OpenCV

What is the best practice in ROS projects when you have to deal with NxM matrices, Nx1 (or Mx1) vectors, sums, multiplications and other basic stuff? Which approach is neater?

I usually use STL structures like std::vector and I'd like to avoid boring cycles to convert data from one type to another, if possibile.

Thank you all.

Asked by alextoind on 2015-09-22 12:55:18 UTC

Comments

Answers

I don't know if there's a "best" practice, but I use Eigen.

Asked by Tom Moore on 2015-09-23 19:48:23 UTC

Comments

Thanks for the replay. I've made the same choice also because of the Eigen/tf conversions.

Asked by alextoind on 2015-09-24 03:05:57 UTC

Agreed, for matrices in C++, Eigen has been #1 for a while and looks like it will stay that way. Not only is it very fast, it is very neat and easy to use.

Asked by Airuno2L on 2015-09-24 07:01:48 UTC