What makes an automotive display look cluttered? integrating objective visual metrics and human perception

Large pillar-to-pillar digital dashboards are increasingly common in modern vehicles, integrating navigation, vehicle status, infotainment, and driver assistance information within a single visual interface. The visual richness of these displays may introduce perceptual clutter that impairs information retrieval and driver safety. This study examined how visual properties of large digital automotive displays relate to perceived clutter and visual search performance using a multi-method approach combining subjective ratings, objective computational metrics, and behavioral measures. Fifty-five participants viewed 18 AI-generated dashboard prototype videos, performed a visual detection task, and rated displays on clutter, density, overlap, colorfulness, and brightness using unipolar rating scales. Objective visual metrics—dynamic entropy, edge density, brightness, colorfulness, and overlap—were computed from video frames. Linear mixed-effects models, multilevel mediation analysis, and exploratory factor analysis were applied. Subjective density and overlap were the strongest predictors of perceived clutter, consistent with spatial complexity theory. Objective entropy and brightness significantly predicted clutter perception, while objective metrics showed close alignment with corresponding subjective ratings. Overlap, edge density, and colorfulness predicted visual search time. Mediation analysis demonstrated that perceived clutter partially mediated the relationship between objective visual properties and search performance. Exploratory factor analysis identified two latent perceptual dimensions: spatial complexity (density, overlap) and surface appearance (colorfulness, brightness). These findings provide empirical support for the primacy of spatial organization in automotive display clutter and offer actionable design guidelines for legible, safe wide-format cockpit interfaces.