Object Detection Model for Aruco Markers


We need to detect the position of fiducial markers in an outdoor setting (1-20m approx distance).

We have tried ArUco marker detection with openCV but get poor distance results that do not take into account the zed stereo vision distance. This could be improved with some manual intervention.

However we wish to know if there is a good open source AI model for detecting fiducial markers that we could use with zed object detection using its custom detector API? It would be great to get something that is more robust and accurate than our current ArUco implementation if possible.

We looked for a model online we no success but assume that this is a common use case and probably exists somewhere (and is publicly available).

Hi @Blake-James
have you tried with our zed-aruco repository?