Markers CCTAG with Python.

[AdonaiVera]

Hi Stella Vslam Team,
First, thank you for this fantastic project; I'm trying to use CCTAG to scale my point cloud. I know that there is already an implementation with AURco; However, I need to use CCTAG; the process that I'm doing in Python is the next one:

1. I use the .map file to get 3d points, rotation, and translation matrix from camera to world and vice-versa.
2. I found the marker on the images, and I get the X and Y of the equirectangular image.
3. I'm transforming from Pixel to bearing.
4. Then I have to transform to coordinate worlds. (I'm trying to apply some Rigidly (+scale) to align two point clouds with known point-to-point correspondences with least-squares error. (I have the pixels of the 3d points that I get from opevslam, and I can create a new rotation and translation matrix using this relation)
5. I take the bearing and found the marker in the world coordinate.

My results are inaccurate; I'm trying to understand how you find the marker in the world coordinate system; any advice would be excellent.
My issue is in step four; I believe there is another way to go from bearing to the world coordinate.

Thank you

[ymd-stella]

What is Rigidly?

[AdonaiVera]

Rigid transformation

[ymd-stella]

It is recommended to inherit marker_detector::base and add marker_detector::cctag.

(Not sure what you mean by "with Python.")

[AdonaiVera]

I'm creating all the modules in Python. I only use the msg file that I generated using Stella vslam. However, I'll execute it first in c++, as you suggest.
Thank you