AprilTag: A robust and flexible visual fiducial system

Abstract

Abstract-While the use of naturally-occurring features is a central focus of machine perception, artificial features (fiducials) play an important role in creating controllable experiments, ground truthing, and in simplifying the development of systems where perception is not the central objective.We describe a new visual fiducial system that uses a 2D bar code style "tag", allowing full 6 DOF localization of features from a single image. Our system improves upon previous systems, incorporating a fast and robust line detection system, a stronger digital coding system, and greater robustness to occlusion, warping, and lens distortion. While similar in concept to the ARTag system, our method is fully open and the algorithms are documented in detail.

Automated indoor RFID inventorying using a self-guided micro-aerial vehicle

“…To evaluate the accuracy of the proposed methods, ground truth position and orientation estimates have been obtained using an external camera to track a fiducial marker attached to the quadrotor. The fiducial marker tracking method, referred to as AprilTags [12], is capable of providing highly accurate, full six degree of freedom (6DOF) pose estimates in real time. Figure 2 shows the AprilTag used for visual tracking mounted to the top of the quadrotor.…”

Towards Autonomous Robotic In-Situ Assembly on Unstructured Construction Sites Using Monocular Vision

“…This requirement has posed a significant challenge for ARC because even by using current state-of-the-art simultaneous localization and mapping (SLAM) techniques, such accuracy is hard to achieve at large scales [12]. In order to address this issue, the authors chose to use computer-vision-based pose estimation algorithms that can achieve high accuracy locally around a visual marker [13,14].…”

“…As pointed out in [11], this requires the localization accuracy to be at least at centimeter level, which is far from achievable using state-of-the-art SLAM style techniques for mobile robots. In order to address this challenge, the authors propose a convenient and accurate solution using planar marker-based pose estimation [14,13], as shown in Figure 6. By 1) attaching fiducial markers at appropriate locations on-site where building tasks are to be performed, 2) surveying their poses m b T in the building reference frame using a total station, and 3) storing these poses inside the system's database, a mobile robot can readily estimate its base's pose b r T inside the building reference frame using equation (3) whenever its on-board camera detects such a marker, based on previous calibration results:…”

“…This 3D rig was made by attaching 18 Apriltags [14] on two intersecting planes forming a 90 degree angle (as shown in the top of Figure 3) so that the 3D coordinate X of each Apriltag's center could be readily measured. The process of calibration was then simply taking a sequence of N images of the rig and inputting them into the author-developed camera calibration tool 4 , which takes advantage of the Apriltag detection algorithm to detect the 2D image coordinate U of each tag center and establish correspondence with X.…”

Duckietown: An Innovative Way to Teach Autonomy

“…All three sign types utilize AprilTags [10] to convey information to the Duckiebots. The use of AprilTags in unison to images/text allows for a progression of functionality and programming skill sets.…”