A nice example of AprilTags that I came across:
We used computer vision and fiducial makers for pose estimation of the sample holder. … The fiducial markers were from the marker family DICT_APRILTAG_36h11 in the ArUco class of markers, and were affixed to a common relative location on all samples, and printed to at 26.65 cm in width and height.
Fig. 6 Pose detection live. The robot views the sample library and dynamically detects the sample data and coordinate reference frames for the sample holders (top). This information can be added to the RViz visualization (bottom).
Xref: Is there a place I can buy a large set of AprilTags of various sizes?