Navigational Pedestrian Movement Model with Vision-driven Agents

Abstract

This paper proposes a navigational pedestrian movement model based on vision-driven agents. In the paper, two different patterns of movement are discussed: natural movement and navigational movement. The natural movement model, which is based on Gibson′s ecological theory of perception and Hillier′s research, facilitates understanding of the relation between visual perception and movement; unfortunately, it has no origin-destination (OD) pairs. In contrast, the navigational movement model does not take the process of visual perception into account. The proposed navigational pedestrian movement model integrates these two movement models by replacing an attractive factor in the conventional natural movement model with a potential field of walls for navigational movement. Consequently, the results of simulations performed show that the proposed model generates a smoother and more plausible movement pattern that is affected by the visual field angle of agents. The problem and its solution according to visual field angle are also demonstrated through model tests. The proposed model demonstrates that an integrated approach can enrich pedestrian studies and enable understanding of the relation between visual perception and movement. It may also be developed as a simulation model for emergencies or wayfinding.