A high-precision crosstalk coefficient calibration method based on maximum distance rotating fringes and noise suppression by spatial matching filtering

Channel crosstalk is a significant factor affecting the measurement accuracy of the color fringe projection method. An accurate crosstalk coefficient calibration algorithm is crucial for the signal separation of color channels. Existing calibration algorithms often suffer from poor precision, complex procedures, and stringent environmental requirements. To address these shortcomings, we propose a new crosstalk coefficient calibration method. Compared with classical and newly proposed calibration methods, our method improves calibration efficiency by at least 83% and performance by at least 70%. The core innovation of our method lies in the introduction of maximum distance rotating fringes (MDRF), which ensure maximum frequency domain separability of signals from different channels. This allows us to estimate the crosstalk coefficient by projecting just one fringe pattern. By leveraging the fact that the primary signal of each channel is significantly larger than the crosstalk signal, we construct a spatial matching filter to maximize the signal-to-noise ratio (SNR) of the crosstalk coefficient estimator. This effectively suppresses the impact of background and quantization noise on estimation accuracy. •We designed maximum distance rotating fringes (MDRF) for optimal frequency separability.•MDRF enables crosstalk coefficient estimation with just one fringe pattern projection.•We developed a spatial matching filter to maximize SNR, boosting estimation accuracy.•Our method enhances calibration efficiency by at least 83% and performance by at least 70%.