A user preference modelling method for the assessment of visual complexity in building façade

This work aims to provide a method to assess the perceptual impact of visual complexity of building facades. The research identifies three variables that incorporate the effect of visual complexity. These variables are the number of design elements and the variation in their position and colour. It introduces the concepts of vertices and corners as atomic indicators on which the physical measurement of the three variables is built. The study designs an experiment, which involves measuring visual complexity and its variables in images of building facades using image-processing techniques, collecting participant's' characteristics and reactions towards the images through an online questionnaire and statistically analysing the relationships between these measures and the reactions and characteristics of participants. The research offers a quantitative lens on the effect of visual complexity as a comprehensive phenomenon on preferences. The study demonstrates that the three variables can be systematically measured, and shows that participants have common visual reactions toward the aspects of visual complexity in images of building facades. This uniformity is confirmed by a regression model, which provides an adequate fit of the three variables as independent variables and preference as a dependent variable. It offers an objective method to assess visual complexity in images of building facades according to common optimal values of the three variables as guidelines to evaluate the design of building facades. These optimal values correspond to the average of the highest rates of the preferences of residents. Planning authorities and design firms can use this method as an objective way to evaluate design alternatives based on the preferences of residents