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.
|Number of pages||13|
|Journal||IEEE Transactions on Intelligent Transportation Systems|
|Publication status||Published - 2017 Feb|
Bibliographical noteFunding Information:
This work was supported in part by the Ministry of Science and Technology under Grant MOST 105-2633-E-002-001, Grant MOST 104-2622-8-002-002, and Grant MOST 104-2627-E-002-001 and in part by National Taiwan University under Grant NTU-ICRP-105R104045. The work of M.-H. Yang was supported in part by the U.S. National Science Foundation through the Faculty Early Career Development (CAREER) Program under Grant 1149783.
All Science Journal Classification (ASJC) codes
- Automotive Engineering
- Mechanical Engineering
- Computer Science Applications