Abstract
Localization is essential for the autonomous navigation of a mobile robot. Because the error of the internal dead-reckoning localization increases with time, the localization error should be corrected periodically using external environmental information. This study presents a practical application of Quick Response (QR) codes to correct the localization error by combining the external QR code information and the internal encoder values. Each QR code contains the information of its own global coordinates and is attached as a one-dimensional array on vertical, thus walls reducing the number of QR codes required for localization. Whenever a QR code is detected and decoded by the mobile robot, the relative transformation between the QR code and the robot can be computed and the pose estimation for the robot is corrected by the extended Kalman filter. Experimental results demonstrate the performance of the localization and autonomous navigation of the mobile robot. The influence of the distance between adjacent QR codes in the environment on the localization performance was also addressed and verified by experiments.
| Original language | English |
|---|---|
| Pages (from-to) | 28384-28390 |
| Number of pages | 7 |
| Journal | IEEE Access |
| Volume | 11 |
| DOIs | |
| State | Published - 2023 |
Keywords
- artificial landmark
- autonomous mobile robot
- extended kalman filter
- Localization
- QR code