Fringe calibration using neural network signal mapping for structured light profilometers

We present a novel neural network signal calibration technique to improve the performance of triangulation based structured light profilometers. The performance of such profilometers is often hindered by the capture of noisy and aberrated pattern intensity distributions. We address this problem by employing neural networks and a spatial digital filter in a signal mapping approach. The performance of the calibration technique is gauged through both simulation and experimentation, with simulation results indicating that accuracy can be improved by more than 80%.