Real-Time Vehicle Localization using on-Board Visual SLAM for Detection and Tracking

Nabil Belbachir, Nadia Noori, Benyamin Akdemir; Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2019, pp. 0-0

Abstract


This paper presents an on-board system based using a monocular vision system, auxiliary sensors and a visual SLAM method for real-time vehicle localization and tracking in natural outdoor environments. A monocular panamorph camera providing 360deg panoramic views has been integrated with a gyro, compass and accelerometer to form a sensor system for installation on-board vehicles. A direct method was applied for detection and tracking of environment based on the images and combined with a calibration of the keypoints using the auxiliary sensors for providing an accurate trajectory to the vehicle. An-board computer is used to facilitate a standalone deployment onboard vehicle. The system was developed and designed as an integral component of a security monitoring and tracking system for humanitarian missions and convoys operating in regions with poor or absence of GPS signal coverage. Preliminary results show the applicability of the system with acceptable accuracy compared to GPS after evaluation in natural outdoor environments in Europe.

Related Material


[pdf]
[bibtex]
@InProceedings{Belbachir_2019_ICCV,
author = {Belbachir, Nabil and Noori, Nadia and Akdemir, Benyamin},
title = {Real-Time Vehicle Localization using on-Board Visual SLAM for Detection and Tracking},
booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops},
month = {Oct},
year = {2019}
}