Embodied AI Glossary中文

ArUco Marker

ArUco码Advanced

A square tag with a black border around a black-and-white grid code; a camera that spots it can compute its own pose relative to the tag.

ArUco is a square fiducial marker system and open-source detection library developed by the AVA research group (Rafael Muñoz, Sergio Garrido, and others) at the University of Córdoba in Spain. Each marker has a thick black border around a binary matrix of black-and-white cells encoding an ID: the black border makes it fast to find in an image, and the binary code allows error detection and correction, as well as telling how far the marker has rotated. A usable set of codes is called a dictionary — a 4×4 marker, for instance, encodes 16 bits. The four corners of a single marker are enough to solve for the camera’s 6D pose relative to the marker using PnP. OpenCV’s aruco module is built on this library, and from version 4.7 it moved into the objdetect module. In robotics, ArUco markers are commonly used for hand-eye calibration, ChArUco calibration boards (ArUco markers embedded in a checkerboard), or stuck on objects and worktables to quickly get ground-truth pose. It belongs to the same family of techniques as AprilTag.

ExampleWhen collecting data, sticking an ArUco marker to a table corner lets the wrist camera detect it every frame and convert the camera pose into the table’s coordinate frame, which also serves as a quick check on how accurate the hand-eye calibration is.

Also called
ArUco, Fiducial Marker
Related
AprilTag · Calibration Board · Hand-Eye Calibration · Perspective-n-Point · Camera Calibration · OpenCV (Open Source Computer Vision Library)
Sources
OpenCV 教程:Detection of ArUco Markers (Chinese)
ArUco 项目页(University of Córdoba, AVA group) (Chinese)

See it in the full glossary →