What about OpenMV’s own computing capability? Can feature point recognition by visual acquisition achieve self-positioning?
Hi, You can achieve this using AprilTags.
See:
http://docs.openmv.io/library/omv.image.html?highlight=april#image.apriltag.x_translation
and a calibrated camera (I calibrated mine by saving images a series of chessboard images and using OpenCV offline.)
Ed
Do you mean something like SLAM ? no, that requires much more memory and processing power than the H7.