Currently, I am using two cameras to estimate the pose of an object (CharucoBoard).
I have tried solvePnP to get the pose of the CharucoBoard using one camera. However, there is still a linear trend of the back-front translation axis. Therefore, I want to use two cameras in a stereo-setup to achieve higher accuracy.
I still don't know how to implement this idea. My guess is to use the triangulatePoints function to get the 3D world coordinates of the object, and then re-input them into solvePnP algorithm, should it work or is there any other suggestion?
