Cost-effective camera based ground truth for indoor localization

Abstract

One of the key requirements for the evaluation of indoor localization systems is an accurate and reliable ground truth. Existing ground truth systems are often expensive due to high hardware cost and complex deployment. In this work, we present a simple yet highly accurate approach for a cost-effective ground truth system based on off-the-shelf infrastructure cameras and printable markers. We developed a marker detection algorithm and systematic 3-layer projection approach between multiple coordinate systems which achieves a median accuracy of 0.48cm, 0.05 degrees and a minimum accuracy of 0.75cm, 0.27 degrees for 2D position and orientation

A Rapid Prototyping Environment for Cooperative Advanced Driver Assistance Systems

“…We therefore have validated the combined dynamics with a maximum speed of around 25 km/h in an oval of 24 × 11.5 meter. For position tracking, we used the camera-based localization system described in [68]. Due to the size of the area to be covered for the oval maneuver, the camera was positioned at a height of 21.6 meters at the south side of the area.…”

“…Due to the size of the area to be covered for the oval maneuver, the camera was positioned at a height of 21.6 meters at the south side of the area. The lateral viewing angle and height lead to a calculated precision between 6.6 cm and 4.8 cm depending on the distance between the tracked object and the camera [68]. The experimental data is split at the points in time when the vehicle starts and stops at the equilibrium point.…”