Fig. 13: Flowchart of deep-learning-based color FPP and the 3D reconstruction results of different approaches.

a The flowchart of deep-learning-based color FPP: CNN predicts the sine and cosine terms related to high-quality wrapped phase map from the input 3-channel fringe images of different frequencies, as well as a “coarse” absolute phase map. Then the outputs of CNN are used to obtain a high-accuracy absolute phase for further 3D reconstruction. b The input color fringe pattern of a David plaster model. c The 3D reconstruction result of the color-coded approach proposed by Zhang et al.³⁰¹. d The 3D reconstruction result of our deep-learning-based method. e Ground truth. f One frame of the color fringe patterns of a 360° rotated workpiece. g 3D result of (f). h, f Registration results viewed from two different perspectives. a–i Adapted with permission from ref. ³⁰⁰, Optica Publishing