An artificial landmark-based indoor navigation system was developed to provide reliable and high accuracy indoor navigation service to an autonomous Unmanned Aerial Vehicle (UAV) with a low computational cost. In this approach, the position of the UAV with respect to the indoor environment was determined by the position information extracted from the artificial landmarks. The artificial landmarks were constructed based on QR codes which stored a unique text string that refer to an absolute coordinate of the indoor environment. Machine learning approaches were applied to provide the system with the ability to detect the landmark through the vision streamed by the aircraft. The system has been experimentally evaluated for proof of concept under practical conditions. The experiment results have proved that using QR code-based artificial landmark as a low computational cost solution is possible to achieve reliable and high accuracy indoor navigation service for an autonomous UAV.
[1]
Trevor Darrell,et al.
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
,
2013,
2014 IEEE Conference on Computer Vision and Pattern Recognition.
[2]
Horst Hellbrück,et al.
QRPos: Indoor positioning system for self-balancing robots based on QR codes
,
2016,
2016 International Conference on Indoor Positioning and Indoor Navigation (IPIN).
[3]
Kaiming He,et al.
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
,
2015,
IEEE Transactions on Pattern Analysis and Machine Intelligence.
[4]
Wei Liu,et al.
SSD: Single Shot MultiBox Detector
,
2015,
ECCV.
[5]
De Xu,et al.
Ceiling-Based Visual Positioning for an Indoor Mobile Robot With Monocular Vision
,
2009,
IEEE Transactions on Industrial Electronics.