Phase-shifting profilometry combined with Gray-code patterns projection: unwrapping error removal by an adaptive median filter

Zheng, Dongliang; Da, Feipeng; Kemao, Qian; Seah, Hock Soon

doi:10.1364/oe.25.004700

Cited by 142 publications

(51 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Field images typically contain various sources of noise, which will affect the final training results. In order to remove highfrequency noise from the images, a Robert detection operator was applied to extract the edge of the broccoli image (Chaudhuri and Chanda, 1984), followed by a median filter with a size of 3 * 3 pixels to remove the noise from the images (according to the size of broccoli head and flower bud displayed in the images) (Zheng et al, 2017).…”

Section: Background Denoisingmentioning

confidence: 99%

A Monitoring System for the Segmentation and Grading of Broccoli Head Based on Deep Learning and Neural Networks

Zhou

et al. 2020

Front. Plant Sci.

View full text Add to dashboard Cite

Section: Background Denoisingmentioning

confidence: 99%

A Monitoring System for the Segmentation and Grading of Broccoli Head Based on Deep Learning and Neural Networks

Zhou

et al. 2020

Front. Plant Sci.

View full text Add to dashboard Cite

“…phases is significantly reduced by using DL-TPU. For these low SNR region, the remaining phase errors have the characteristics of accumulation and can be easily further corrected by some compensation algorithm for fringe order errors [42]- [44] (refer to Supplementary Section 4 for details of these compensation algorithms). Consequently, the trained models can substantially decrease error points to provide better phase unwrapping results (even f h = 64) and lower error rates, which demonstrates the capability and reliability of DL-TPU for phase unwrapping.…”

Section: A Quantitative Comparison With Mf-tpumentioning

confidence: 99%

Temporal phase unwrapping using deep learning

Yin

Chen

Feng

et al. 2019

Sci Rep

113

View full text Add to dashboard Cite

The multi-frequency temporal phase unwrapping (MF-TPU) method, as a classical phase unwrapping algorithm for fringe projection techniques, has the ability to eliminate the phase ambiguities even while measuring spatially isolated scenes or the objects with discontinuous surfaces. For the simplest and most efficient case in MF-TPU, two groups of phase-shifting fringe patterns with different frequencies are used: the high-frequency one is applied for 3D reconstruction of the tested object and the unit-frequency one is used to assist phase unwrapping for the wrapped phase with high frequency. The final measurement precision or sensitivity is determined by the number of fringes used within the high-frequency pattern, under the precondition that its absolute phase can be successfully recovered without any fringe order errors. However, due to the non-negligible noises and other error sources in actual measurement, the frequency of the high-frequency fringes is generally restricted to about 16, resulting in limited measurement accuracy. On the other hand, using additional intermediate sets of fringe patterns can unwrap the phase with higher frequency, but at the expense of a prolonged pattern sequence. With recent developments and advancements of machine learning for computer vision and computational imaging, it can be demonstrated in this work that deep learning techniques can automatically realize TPU through supervised learning, as called deep learning-based temporal phase unwrapping (DL-TPU), which can substantially improve the unwrapping reliability compared with MF-TPU even under different types of error sources, e.g., intensity noise, low fringe modulation, projector nonlinearity, and motion artifacts. Furthermore, as far as we know, our method was demonstrated experimentally that the high-frequency phase with 64 periods can be directly and reliably unwrapped from one unit-frequency phase using DL-TPU. These results highlight that challenging issues in optical metrology can be potentially overcome through machine learning, opening new avenues to design powerful and extremely accurate high-speed 3D imaging systems ubiquitous in nowadays science, industry, and multimedia.

show abstract

“…From (5), the recovered fringe order map ̂ (with entries ̂( , )) is the superposition of the true order map (with entries ( , )) and the errors (with entries ( , )). We propose to separate these two components so that we are able to recover the correct fringe order map.…”

Section: Fringe Order Errors In Temporal Phase Unwrappingmentioning

confidence: 99%

“…In order to eliminate the phase ambiguity, phase unwrapping is required to recover the absolute phase maps (which are continuous). Among many phase unwrapping methods, temporal phase unwrapping is commonly used, especially for applications with noise and discontinuities [3][4][5][6].…”

Section: Introductionmentioning

confidence: 99%

A new method for fringe order error correction in fringe projection profilometry

Zhang

Duan

et al. 2019

Optical Metrology and Inspection for Industrial Applications VI

View full text Add to dashboard Cite

© 2019 SPIE. Phase unwrapping is an essential step for 3D shape measurement based on fringe projection. Temporal phase unwrapping methods can be implemented by analyzing the multiple patterns that encode the fringe order information. They can retrieve the fringe orders on a pixel-by-pixel basis and are less prone to error propagation compared with spatial methods. However, fringe orders errors may still occur due to noise, reflectivity fluctuation and discontinuity of the object surface. Such errors may exhibit an impulsive nature and result in significant error to the recovered absolute phase map. This has been exploited by several methods to correct the fringe order errors, e.g., by filtering the fringe order sequences in a line-by-line manner. In this paper, a new method is proposed to correct the errors associated with fringe orders for the temporal phase unwrapping. The scheme first makes use of the lowrankness property of the fringe order map and sparse nature of the impulsive fringe order errors to more effectively remove the impulsive errors by applying robust principal component analysis (RPCA) algorithm. Then the smoothness of the two-dimensional unwrapped phase map is examined and the residual fringe order errors are detected based on a discontinuity measure of the phase map and corrected by comparing phase difference between two adjacent pixels in the unwrapped phase. The effectiveness of the proposed method is demonstrated via numerical experiments.

show abstract

Phase-shifting profilometry combined with Gray-code patterns projection: unwrapping error removal by an adaptive median filter

Cited by 142 publications

References 31 publications

A Monitoring System for the Segmentation and Grading of Broccoli Head Based on Deep Learning and Neural Networks

A Monitoring System for the Segmentation and Grading of Broccoli Head Based on Deep Learning and Neural Networks

Temporal phase unwrapping using deep learning

A new method for fringe order error correction in fringe projection profilometry

Contact Info

Product

Resources

About