Adaptive microphase measuring profilometry for three-dimensional shape reconstruction of a shiny surface

Abstract. Three-dimensional (3-D) reconstruction of a shiny surface is a great challenge for existing structured light techniques because serious measurement errors will occur in image saturation regions in the captured image. To this end, an adaptive microphase measuring profilometry (AMPMP) is proposed. In this method, an optimal frequency selection strategy is introduced to select the frequencies of phase-shifted fringe patterns. The intensity relation between the projected pattern and the captured image is estimated by fitting a nonlinear function after projecting a series of uniform gray-level patterns onto the shiny surface, and the mapping regions of saturated areas in the captured image are computed with the correspondence between the projector image plane and the camera image plane, which is established through MPMP. Based on the estimated intensity relation and the mapping regions, the optimal projection intensities of saturated areas are adaptively adjusted. Then, the adjusted phase-shifted fringe patterns are projected to obtain the absolute phase distribution of the shiny surface. Therefore, accurate 3-D shape reconstruction is realized. The experimental results demonstrated that the proposed AMPMP can well tackle saturated areas of shiny surfaces and has greater performance than other commonly used adaptive fringe projection techniques.