Large depth-of-field fringe projection profilometry applied to defocused scenes.

Abstract

Fringe projection profilometry (FPP) is a widely adopted technique for three-dimensional (3D) reconstruction. However, its depth-of-field (DOF) is constrained when reconstructing defocused scenes, mainly due to limitations in the camera model and image blur. This study introduces a camera model based on the ideal optical system, which effectively reduces the systematic errors associated with the conventional pinhole camera model. A calibration method to determine the optical system parameters of the improved camera model is proposed. Additionally, the point spread function (PSF) of the camera is calibrated and the image is deblurred through non-blind deconvolution, thereby minimizing the phase aliasing resulting from defocus. Experimental results validate the potential of the proposed method for accurate 3D reconstruction in scenes with a wide depth range.