Visual complexity and gaze behavior in urban and forest scenes: Evidence from eye-tracking and image features

Abstract

Viewing natural scenes has been shown to produce fewer but longer eye fixations than viewing urban scenes, a pattern often interpreted within Attention Restoration Theory (ART) as reflecting effortless attention and soft fascination. However, the perceptual mechanisms underlying these differences remain unclear. This study examined whether gaze behavior can be explained by objective and subjective measures of visual complexity within the framework of ART. Sixty-four participants viewed photographs of urban façades and deciduous forests while their eye movements were recorded. Objective visual complexity was quantified using entropy, edge density, and fractal complexity, while perceived complexity and liking were assessed in an independent sample. Urban scenes elicited significantly more fixations and shorter fixation durations than forest scenes. However, when objective image characteristics were included in multivariate multiple regression models, scene category was no longer a significant predictor of either fixation measure. Instead, fractal complexity significantly predicted a greater number of fixations. In models including subjective evaluations, higher perceived complexity predicted shorter fixation durations, while greater liking predicted fewer fixations. These findings suggest that differences in visual attention between natural and urban environments cannot be attributed solely to environmental category and may also reflect specific structural and perceptual characteristics of visual scenes.