Color differences in complex scenes: the role of background

Color appearance is strongly shaped by surrounding context, yet color-quality assessments for multicolored materials, such as camouflage textiles, typically rely on judgments made against uniform backgrounds. This study systematically tests whether a homogeneous background matched to the average color of a multicolored surround can approximate the perceptual influence of the original patterned background when assessing small color differences in MARPAT camouflage. Using a center-background paradigm, pairs of stimuli with either identical centers or small lightness differences (<i>Δ</i><i>L</i>^∗=±0.5, ±1.0) were presented on three background types: a MARPAT two-color checkerboard, its corresponding CIELAB-matched average background, and a neutral gray reference. Psychophysical experiments with 20 color-normal observers revealed systematic perceptual discrepancies between patterned and color-averaged backgrounds: even physically identical centers produced mean visual differences of ∼0.2-1.2<i>Δ</i><i>E</i>₀₀ (1:1:1). Background composition exerted a strong effect (<i>p</i><0.001), whereas checkerboard size (4×4, 8×8, 16×16) had no significant influence. The largest induced differences occurred for combinations in which the center-surround lightness ratio approached unity, consistent with enhanced chromatic induction. Sensitivity to small <i>Δ</i><i>L</i>^∗ differences declined sharply when stimuli were spatially separated, with many observers failing to detect even ±1.0<i>Δ</i><i>L</i>^∗ differences on gray. Substantial intra- and inter-observer variability further highlighted the difficulty of judging subtle differences in complex scenes. Baseline measurements on uniform gray backgrounds yielded near-zero visual differences, confirming that the large apparent differences observed under patterned surrounds were induction-driven rather than due to random noise. Overall, the results demonstrate that a color-averaged background does not serve as a perceptually equivalent substitute for a complex multicolored surround, underscoring the need for context-appropriate backgrounds in the visual evaluation of camouflage, textiles, displays, and other patterned materials.