This paper presents an algorithm for ego-positioning by using a low-cost monocular camera for systems based on the Internet-of-Vehicles. To reduce the computational and memory requirements, as well as the communication load, we tackle the model compression task as a weighted k-cover problem for better preserving the critical structures. For real-world vision-based positioning applications, we consider the issue of large scene changes and introduce a model update algorithm to address this problem. A large positioning data set containing data collected for more than a month, 106 sessions, and 14 275 images is constructed. Extensive experimental results show that submeter accuracy can be achieved by the proposed ego-positioning algorithm, which outperforms existing vision-based approaches.
Vision-Based Positioning for Internet-of-Vehicles
2017
Article (Journal)
English
Vision-Based Positioning for Internet-of-Vehicles
IEEE | 2017
|European Patent Office | 2021
|Vehicle positioning method and system of Internet of Vehicles
European Patent Office | 2022
|European Patent Office | 2015
|