Defocused binary fringe phase error modeling and compensation using depth-discrete Fourier series fitting.

The binary defocus fringe projection is a widely adopted way to increase the speed of fringe projection profilemotry. However, the projected patterns may deviate from ideal ones at some depths. We propose a theoretical model and a corresponding compensation method to explain and calibrate the phase error of defocus-projected patterns. We first low-pass filter the projected patterns at different depths to obtain corresponding ideal ones. Then, we calibrate the model coefficients based on the errors between the original and ideal fringe patterns. The calibrated phase error model can be used to compensate the phase error at arbitrary depths within the calibration volume. Experiments are conducted to verify the feasibility and performance of the proposed method.