Wavelet based deep learning for depth estimation from single fringe pattern of fringe projection profilometry

Zhu, Xinjun; Han, Zhiqiang; Song, Limei; Wang, Hongyi; Wu, Zhichao

doi:10.1007/s11801-022-2082-x

Cited by 7 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep learning-based approaches have been adopted in optical measurement and experimental mechanics to accomplish several classic tasks such as phase extraction, fringe analysis, interferogram denoising, and deformation determination [32][33][34][35][36][37]. Numerous devoted topics in 3D shape measurement involving deep learning have been introduced in the last few years.…”

Section: Introductionmentioning

confidence: 99%

Generalized Fringe-to-Phase Framework for Single-Shot 3D Reconstruction Integrating Structured Light with Deep Learning

Nguyen

Lam

et al. 2023

Sensors

View full text Add to dashboard Cite

Three-dimensional (3D) shape acquisition of objects from a single-shot image has been highly demanded by numerous applications in many fields, such as medical imaging, robotic navigation, virtual reality, and product in-line inspection. This paper presents a robust 3D shape reconstruction approach integrating a structured-light technique with a deep learning-based artificial neural network. The proposed approach employs a single-input dual-output network capable of transforming a single structured-light image into two intermediate outputs of multiple phase-shifted fringe patterns and a coarse phase map, through which the unwrapped true phase distributions containing the depth information of the imaging target can be accurately determined for subsequent 3D reconstruction process. A conventional fringe projection technique is employed to prepare the ground-truth training labels, and part of its classic algorithm is adopted to preserve the accuracy of the 3D reconstruction. Numerous experiments have been conducted to assess the proposed technique, and its robustness makes it a promising and much-needed tool for scientific research and engineering applications.

show abstract

Section: Introductionmentioning

confidence: 99%

Generalized Fringe-to-Phase Framework for Single-Shot 3D Reconstruction Integrating Structured Light with Deep Learning

Nguyen

Lam

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Therefore, more advanced algorithms have been developed to extract absolute phases directly from a single-shot, single-frequency fringe pattern. Typical works include the untrained deep learning-based method [12] and wavelet-based deep learning method [13]. The former achieves absolute phase retrieval with two networks, where the first one refines the relative phases and produces a coarse fringe order for unwrapping and the second one unwraps the relative phases with the fringe order and then refines them.…”

Section: Introductionmentioning

confidence: 99%

PSNet: A Deep Learning Model-Based Single-Shot Digital Phase-Shifting Algorithm

Qi,

Liu,

Pang

et al. 2023

Sensors

View full text Add to dashboard Cite

In contrast to traditional phase-shifting (PS) algorithms, which rely on capturing multiple fringe patterns with different phase shifts, digital PS algorithms provide a competitive alternative to relative phase retrieval, which achieves improved efficiency since only one pattern is required for multiple PS pattern generation. Recent deep learning-based algorithms further enhance the retrieved phase quality of complex surfaces with discontinuity, achieving state-of-the-art performance. However, since much attention has been paid to understanding image intensity mapping, such as supervision via fringe intensity loss, global temporal dependency between patterns is often ignored, which leaves room for further improvement. In this paper, we propose a deep learning model-based digital PS algorithm, termed PSNet. A loss combining both local and global temporal information among the generated fringe patterns has been constructed, which forces the model to learn inter-frame dependency between adjacent patterns, and hence leads to the improved accuracy of PS pattern generation and the associated phase retrieval. Both simulation and real-world experimental results have demonstrated the efficacy and improvement of the proposed algorithm against the state of the art.

show abstract