Applications for street-view data from moving camera in transportation
Project aims to present novel applications for data collected from moving camera image sequences in transportation.
In recent years, there has been an increased need for high-quality data in transportation to be able to perform efficient management strategies and decisions. To provide high-quality data, video systems are increasingly used for various applications, for example, traffic flow/pedestrian flow analysis from surveillance cameras, spatial-temporal space utilization in urban spaces, passengers flow into public transport and others. Most of
the above-mentioned applications use a static camera (camera placed in a fixed location) and cover a finite area. There is a growing demand for data collection in cities that requires real-time spatiotemporal information from a large expanse of the city. Since continuous and long-term observations for such a large area are practically impossible with just a static camera, there is a potential to use moving cameras for the same.
The project investigates the use of street-view image sequences from a moving monocular camera (e.g., smartphone, dash-cams, or action camera) to collect data for different applications in transport. Currently, the focus is to propose a novel approach for gathering traffic information using these cameras on a road link and estimating traffic states, namely flow, density, and speed. The project also focuses on the investigation of different image processing techniques and algorithms that could be potentially used to collect traffic data.