Accurate 3D reconstruction via fringe-to-phase network.

Learning 3D shape representations from structured-light images for 3D reconstructions has become popular in many fields. This paper presents a new approach integrating a fringe-to-phase network with a fringe projection profilometry (FPP) technique to achieve 3D reconstructions with superior accuracy and speed performance. The proposed fringe-to-phase network has a UNet-like architecture, capable of retrieving three wrapped phase maps directly from a color image comprising three fringe patterns with designated frequencies. Because the phase maps contain the 3D shape representations of the measurement target, they serve as an intermediary to transform the single-shot fringe-pattern image into the 3D shapes of the target. The datasets with ground-truth phase labels are generated by using a tri-frequency FPP method. Unlike the existing techniques, the proposed approach yields both high-accuracy and fast-speed 3D reconstructions. Experiments have been accomplished to validate the proposed technique, which provides a promising tool for numerous scientific research and industrial applications. [Display omitted] • Integration of structured-light technique and deep learning. • Learning 3D shape representation from a single fringe image. • Phase is retrieved directly from a single fringe image. • Yielding higher accuracy than other single-shot techniques. • Providing faster speed than existing accurate measurement techniques. [ABSTRACT FROM AUTHOR]