Sinusoidal fringe-pattern projection for 3-D surface measurement with variable illuminance

This paper presents a method of projecting sinusoidal fringe patterns with modified maximum gray level to accommodate variable ambient illuminance that would otherwise cause intensity saturation and measurement error in phase-shifting surface-shape measurement. The maximum input gray level (MIGL) in the projected patterns can be reduced to an optimal trade-off point, below which the image intensity signal-to-noise ratio would diminish the advantage of further MIGL reduction. Measurement simulations using ten MIGLs (75 to 255) demonstrated reduction in RMS errors for ambient illuminance of 600, 700, 800 and 900 lx, from 0.31, 0.45, 0.75 and 1.21 mm, respectively, to 0.2 mm. The advantage of the approach was confirmed in real measurements of a flat plate and human mask.