Examining the Impact of Color Contrast Using Digital Anthropometric Estimates of Body Composition

This study examined the reliability of adiposity (%Fat) measured from a single digital image using a smartphone-based application with different color backgrounds.

A reference image was obtained while participants (n = 32) stood in front of a white background (WB), with additional images obtained in front of black, green, orange, and gray backgrounds. A contrast ratio was calculated between the background and clothing RGB hexadecimal color values. Linear mixed-effects modeling determined if color contrasts explained the background effect on %Fat estimation.

The Green, Orange, and Gray conditions yielded constant error (CE) roughly 1.0 %Fat, with greater CE observed in the Black background. Neither the within person-contrast (p = 0.638) nor the between-person contrast (p = 0.451) was associated with the CE and did not explain any extra variability in the CE (χ² = 0.83, p = 0.660).

Images with darker background colors may result in greater CE due to lower contrast between the background and clothing.