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

Area of Science:

• Biometrics
• Digital Health
• Image Analysis

Background:

• Smartphone-based applications offer a convenient method for estimating body composition.
• The accuracy of these tools can be influenced by various imaging parameters.
• Background color is a potential factor affecting image analysis for adiposity estimation.

Purpose of the Study:

• To assess the reliability of body fat percentage (%Fat) measurements using smartphone images.
• To investigate the impact of different background colors on %Fat estimation accuracy.

Main Methods:

• Participants (n=32) were photographed against white, black, green, orange, and gray backgrounds.
• A contrast ratio was calculated between background and clothing colors.
• Linear mixed-effects modeling analyzed the effect of color contrast on %Fat estimation error.

Main Results:

• Green, orange, and gray backgrounds resulted in approximately 1.0%Fat constant error.
• Black backgrounds showed a greater constant error.
• Neither within-person nor between-person color contrast explained the error or variability.

Conclusions:

• Darker background colors may increase estimation error in %Fat measurements.
• Lower contrast between clothing and background likely contributes to reduced accuracy.
• Optimizing background selection is crucial for reliable smartphone-based adiposity assessment.