PU-M-Net for phase unwrapping with speckle reduction and structure protection in ESPI

Xu, Min; Tang, Chen; Shen, Yuxin; Hong, Nian; Lei, Zhenkun

doi:10.1016/j.optlaseng.2021.106824

Cited by 19 publications

(12 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Phase unwrapping can be treated as a regression problem in which a neural network directly learns the mapping relationship between the wrapped phase and the absolute phase. [22][23][24][25][26][27][28][29][30][31][32][33] As illustrated in Fig. 4, after being fed with a wrapped phase, the trained network directly outputs the unwrapped (absolute) phase.…”

Section: Deep-learning-performed Regression (Drg) Methodsmentioning

confidence: 99%

“…Zhou et al 31 improved the robustness and efficiency of deep learning phase unwrapping by adding preprocessing and postprocessing. Xu et al 32 improved the accuracy and robustness of phase unwrapping in an end-to-end case by using a composite loss function and adding more skip connections to Res-UNet. Zhou et al 33 used the GAN in InSAR phase unwrapping.…”

Section: Deep-learning-performed Regression (Drg) Methodsmentioning

confidence: 99%

“…• Random matrix enlargement (RME): The absolute phase of various distributions is obtained by enlarging a small random matrix of random size and range. 24,28,32 • Gaussian functions superposition (GFS): The absolute phase of various distributions is obtained by weighted superposition of multiple Gaussian functions with different mean values and variances. 29,35,38,39,[42][43][44] • Zernike polynomials superposition (ZPS): The absolute phase of various distributions is obtained by weighted superposition of Zernike polynomials in different orders.…”

Section: (B)mentioning

confidence: 99%

“…The RME method first generates small square matrices with different sizes (randomly set from 2 × 2 to 8 × 8), different data distribution types (uniform or Gaussian), and then enlarges them into a size of 128 × 128 by different interpolation methods (nearest, bilinear, or bicubic interpolation). 24,28,32 As an example, shown in Fig. 7, an initial small matrix is interpolated and enlarged into a big matrix, which is then linearly mapped to absolute phase with a higher h. In RME, the continuity is guaranteed by interpolation, while the randomness is introduced by the parameter selection of the initial small matrices.…”

Section: Random Matrix Enlargement (Rme)mentioning

confidence: 99%

“…• Deep-learning-performed regression method (dRG) estimates the absolute phase directly from the wrapped phase by the neural network. [22][23][24][25][26][27][28][29][30][31][32][33] The used dataset contains the paired wrapped phase as input and absolute phase as ground truth (GT), as shown in Fig. 2(a).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Deep learning spatial phase unwrapping: a comparative review

Wang

Kemao

et al. 2022

Adv. Photon. Nexus

View full text Add to dashboard Cite

Phase unwrapping is an indispensable step for many optical imaging and metrology techniques. The rapid development of deep learning has brought ideas to phase unwrapping. In the past four years, various phase dataset generation methods and deep-learning-involved spatial phase unwrapping methods have emerged quickly. However, these methods were proposed and analyzed individually, using different strategies, neural networks, and datasets, and applied to different scenarios. It is thus necessary to do a detailed comparison of these deep-learning-involved methods and the traditional methods in the same context. We first divide the phase dataset generation methods into random matrix enlargement, Gauss matrix superposition, and Zernike polynomials superposition, and then divide the deep-learning-involved phase unwrapping methods into deep-learning-performed regression, deep-learning-performed wrap count, and deep-learningassisted denoising. For the phase dataset generation methods, the richness of the datasets and the generalization capabilities of the trained networks are compared in detail. In addition, the deep-learning-involved methods are analyzed and compared with the traditional methods in ideal, noisy, discontinuous, and aliasing cases. Finally, we give suggestions on the best methods for different situations and propose the potential development direction for the dataset generation method, neural network structure, generalization ability enhancement, and neural network training strategy for the deep-learning-involved spatial phase unwrapping methods.

show abstract