Novel methods for analysing pedestrian behaviour in urban areas using moving cameras

Within urban planning, there is a need for improved decision-support using data with real-time spatiotemporal information of pedestrian behaviour. Recent advances within computational urban analytics involving sensing techniques and image processing algorithms show promise to respond to this demand. However, most current applications in the field use a static camera, which does not allow for observations over larger areas.

Addressing this gap in the literature, the aim of this research project is to support urban design using street-view image sequences collected with a moving camera to generate pedestrian count data and heat maps. This will contribute to better understanding of human movement, which is central to urban planning and research. For example, it enables the identification of suitable locations for services such as bike stands, elevators, dustbins, and traffic signs. It also makes it possible to investigate features in the built environment that encourage (or inhibit) walking, such as the presence of greenery.

Previous research has shown that urban design can promote walking, which has numerous advantages over motorised travel related to cost, congestion, health, traffic safety, environmental impact, and social cohesion. Walking-friendly urban planning therefore contributes to reach the global goals in Agenda 2030.

Start date: 15 April 2023

End date: 31 December 2026