High-quality 3D shape measurement by kernel-optimized high sinusoidal similarity dither patterns

The error-diffusion dithering technique is widely used in three-dimensional (3D) measurement. The sinusoid similarity of dithering fringe patterns greatly affects measurement accuracy, while the traditional dithering methods lack this quality. In this paper, we present a kernel-optimized dithering technique based on the Floyd–Steinberg dithering method and establish an optimization strategy considering both intensity and phase aspects to generate high-quality sinusoidal fringes for 3D measurement by utilizing the binary defocusing technique. Both simulations and experiments demonstrate that the proposed method could improve the performance of 3D measurement effectively under different defocusing amounts, especially in narrow strips.