MODELOS ESTADÍSTICOS PARA LA CARACTERIZACIÓN DE LAS PREFERENCIAS POR LOS ATRIBUTOS ESTÉTICOS CONSTITUYENTES DE LOS PAISAJES

Abstract

In the present work a strategy to objectively characterize landscapes based on aesthetics attributes (lines, forms, texture, color, spatial composition and scale) through photographs has been deduced. A set of variables based on physical dimensions (length, angle, area or luminosity) measurable by computer aided design software (CAD) have been proposed. In parallel, population preferences for every analyzed aesthetic attribute have been collected through surveys. Thus, using collected preferences and the deduced set of variables to characterize the aesthetic attributes, respective models to characterized preferences for every attribute have been developed through linear and non-linear regression techniques. Results show that color (R2c=0.522), followed by lines (R2c=0.346) and forms (R2c=0.335), are the most important attributes in order to explain landscape preferences. Accordingly, the deduced expressions to characterize preferences for every attribute, allow incorporating in engineering projects landscape integration criteria according to preferences.