Visualising the structure of architectural open spaces based on shape analysis

Visualisation and interpretation of architectural spaces are intellectually challenging and multifaceted exercises with fairly wide ranging applications and implications. In the context of urban planning, they are most commonly undertaken to evaluate the usage of the architectural space to ensure efficient navigation and accessibility [1]. These exercises clearly assume a certain influence of the built structures on the human cognition. However, what aspect of architectural space affects the human behaviour still remains an open debate. In this respect, it is closely similar to an exercise to identify the unknown visual variables in information visualization. Since a quantitative analysis of the architectural geometric structure on a large scale will be a daunting computational task, the open space bounded by the built structures is studied instead [2]. The study of architectural open spaces essentially involves the computation of the visibility polygon or isovist (space visible all around a viewpoint, see Figure 1a) from a viewpoint and calculating various shape measures of the visibility polygon. The isovist computation involves drawing rays