Abstract

Since GPS signal is not applicable indoors, vehicle tracking has proven a hassle in underground parking structures. Recent solutions highly rely on floor map to constraint inertial sensors noises. In this paper, we propose VeMap, a road map construction system using only smartphones inside vehicles. It saves effort-intensive and time-consuming business negotiations with building operators, and expensive personnel cost to gather such data. It fuses multiple sensors to calibrate inertial noises, and uses Dynamic Time Warping to align multiple trajectories. We represent the floor plan with occupancy grid mapping, and explore a vision-mobile joint algorithm to extract its skeleton and form the road map. VeMap is tested in a 250mx90m parking structure, and it can be directly used for driving navigation to free parking spaces.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call