An improved Fourier transform profilometry and phase coding for 3D shape measurement

Abstract

Optical non-contact 3D shape measurement has attracted wide attention from the scientific community. Although by increasing the frequency of projected fringes, the measurement accuracy can be improved, difficulties in phase unwrapping are induced. Reducing the number of projected fringes can also increase the measurement speed. However, there are still certain challenges to meeting the requirements of high speed and high precision without creating additional projection fringe patterns. Along these lines, an improved 3D shape measurement method based on Fourier transform profilometry and phase coding was proposed in this work, where only four projection fringe patterns must be projected. A sinusoidal grating and a uniform gray-level pattern were used to obtain the wrapped phase recovered by the background-normalized Fourier transform algorithm. Two captured phase coding patterns combine with the uniform gray-level pattern to obtain fringe orders solved by the 2 + 1 phase-shifting algorithm. After the fringe order is calculated, the absolute phase map is retrieved and the 3D shape can be obtained. The proposed method is suitable for isolated and complex objects. The performance of the proposed method in reconstructing the 3D shapes of objects was experimentally verified.