I am currently facing a challenge with my autonomous mobile robot's docking system, which relies on QR code detection (specifically AprilTags) for final positioning and orientation. The core issue arises from the yaw orientation accuracy being adversely affected by the angle at which the onboard camera views the QR code. This perspective distortion leads to inaccuracies in aligning the robot’s position and orientation relative to the docking target.

Problem Description: My system's intended behavior is to navigate the robot to a predetermined goal pose, situated 20 cm in front of the QR code, using the orientation provided by the detected tag. However, as the robot adjusts its alignment during its approach, based on the QR code's orientation, I observe a consequential shift in the goal y-pose. This shift is evidently linked to the camera's viewing angle relative to the QR code, altering the perceived orientation of the tag.

  • $\begingroup$ Are you calibrating your camera? That should help remove some of the effects of the lens distortion. $\endgroup$
    – mjcarroll
    Jan 8 at 15:16


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.