End-to-end Deep Learning Method for Absolute Phase Retrieval

Jin Lu,Chen Yuan,PeiYang Wu,Wei Sun,XiaoGuang Sun,Jian Xu
DOI: https://doi.org/10.1117/12.3007796
2023-01-01
Abstract:Absolute phase plays a crucial role in various applications, including camera or projector calibration, stereo matching, structured light measurement, and fringe projection profilometry (FPP). Recently, significant progress has been made in the development of deep learning-based approaches for absolute phase recovery. Many deep neural networks have been created, improved, or directly integrated into the phase retrieval procedure. Analyzing these methods, a common trend is observed in the sequential calculation of wrapped phase, fringe order, and absolute phase. The accuracy of previous results has a direct impact on the subsequent steps, leading to potential error accumulation and reduced recovery speed. To address these challenges, we propose an end-to-end deep learning method based on Res-UNet that directly predicts the absolute phase from a single fringe image without any additional fringe patterns. The presented approach simplifies the procedure of phase unwrapping and overcomes limitations of existing techniques. To note that, to save cost and labor for training the Res-UNet, a novel and virtual digital fringe project system with 3D Studio Max is also established for generating data close to reality. Experiments have been carried out to validate the performances of the proposed method.
What problem does this paper attempt to address?