Robust and Fast Binary Fringe Projection Profilometry With Error Diffusion Model

Abstract

For scientific and industrial applications requiring high-accuracy and high-speed 3D shape measurement, in this Letter we propose a robust and fast binary fringe projection profilometry (BFPP) method that transcends the traditional paradigm of binary defocusing. First, a perturbed error diffusion algorithm is employed to generate 1-bit quasi-sinusoidal patterns with high accuracy, enabling high-speed 3D measurement since light modulators project 1-bit patterns at a significantly higher speed. Second, differing from traditional methods that rely on intentional projector defocusing to suppress high-order harmonics, which reduces the depth range of measurement and hinders the measurement process, our method can robustly handle both focused and defocused cases. Experimental results show that the proposed method achieves phase and height root-mean-square errors of 0.0012 rad and 0.0884 mm, respectively, demonstrating accuracy comparable to recent binary defocusing methods.