Visual Moment Equilibrium: A Computational Cognitive Model for Assessing Visual Balance in Interface Layout Aesthetics

Quick visual balance perception in layouts is essential for a positive user experience. However, existing computational models often struggle to accurately capture this key aesthetic aspect, particularly in interfaces with asymmetric elements. This paper introduces Visual Moment Equilibrium (VME), a new cognitive model that redefines visual balance as a unified perceptual force field, similar to moment equilibrium in physical systems. Based on principles from Gestalt psychology, spatial cognition, and psychophysics, we incorporate three main innovations: (1) a Measured Balance index enhanced with psychophysical transformations to enable sensitive quantification of visual imbalance; (2) a nine-grid visual weighting system combined with Manhattan distance to reflect human attentional distribution and non-Euclidean spatial reasoning; and (3) a Shape Sparsity Ratio with a piecewise compensation function that formally operationalizes the Gestalt principle of closure, especially for irregular visual elements. Validation against human perceptual benchmarks from the Analytic Hierarchy Process shows that the VME model has a strong correlation with expert judgments regarding regular interfaces (Pearson’s r = 0.942, accounting for 88.8% of the variance), outperforming the widely used model (33.9%). VME also maintains high predictive accuracy for irregular interfaces (r = 0.890), emphasizing its wide applicability across various design configurations.