I have a calibrated (I know the intrinsic parameters) camera. Also, I have 2 photos of the object from this camera from different positions. In addition, I have synchronized data from the lidar of these two photos (distance in meters). A key step is to calculate translation vector in real-world coordinates (in meters).

At the moment I have already received the rotation matrix and translation vector using opencv cv2.findEssentialMat and cv2.recoverPose. And I don't have any ideas how to use depth data to resolve this scale ambiguity.

Thanks!