Deep absolute phase recovery from single-frequency phase map for handheld 3D measurement

Bai, Songlin; Luo, Xiaolong; Xiao, Kun; Song, Wanzhong

doi:10.1016/j.optcom.2022.128008

Cited by 22 publications

(14 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared with the typical semantic segmentation, the similarity between the wrapping phases is high, the channel is single, and the texture changes are not obvious. Therefore, the fringe order of each point in the phase diagram depends more on the object shape, pixel position and other factors 16 . In addition, general deep convolutional neural networks often require complex network structures to achieve good results, which leads to too many parameters and cannot be used on some lowperformance devices.…”

Section: Deep Learning Frameworkmentioning

confidence: 99%

“…To demonstrate the effectiveness of our method and the roles played by different modules in our designed network structure, we conduct ablation experiments on the network. In order to better compare, we use the dataset in Wang et al 's article as the training dataset and validation dataset in this experiment 16 . The results are shown in Table 1.…”

Section: Ablation Experimentsmentioning

confidence: 99%

“…Qin et al proposed a VUR-Net method, which further improved the phase unwrapping performance, but the improvement depended on the expense of significantly increased training cost 15 . Bai et.al presented a learning-based approach of retrieving the absolute fringe order from singlefrequency wrapped phase of tooth model which can operate in realtime for high-resolution phase maps 16 .…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

PEENet for phase unwrapping in fringe projection profilometry

Wang

Zhou

2022

Thirteenth International Conference on Information Optics and Photonics (CIOP 2022)

View full text Add to dashboard Cite

In this paper, we propose a lightweight deep convolution neural network, named PEENet, for high resolution image phase unwrapping in fringe projection profilometry on the device with limited performance. In our method, the dilated convolution strategy is applied to the networks, which increases the receptive field of the network while reducing the amount of network parameters. In the PEENet, we use Atrous Spatial Pyramid Pooling (ASPP) structure which can reduce the network parameters (total 0.48 million) while can extract the deep features of image. We also use Edge-Enhanced Block (EEB) structure, which can enhance the edge features of the image. We conducted ablation experiments to explore the effect of different network structures on network performance and then we compare our method with the other lightweight deep convolution neural network with the same training and testing datasets. We also build a new dataset that contain more different situations which can enhance the generalization ability of the network. The results show that our method achieves higher accuracy with fewer parameters and the new dataset works well.

show abstract

Section: Deep Learning Frameworkmentioning

confidence: 99%

Section: Ablation Experimentsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

PEENet for phase unwrapping in fringe projection profilometry

Wang

Zhou

2022

Thirteenth International Conference on Information Optics and Photonics (CIOP 2022)

View full text Add to dashboard Cite

show abstract

“…1,12,13 Recently, deep learning-based phase unwrapping methods have attracted more and more attention through end-to-end learning from wrapped phase to fringe order to achieve phase unwrapping. [14][15][16][17][18] Essentially, phase unwrapping is to determine the fringe order for each point in the wrapped phase, which can be formulated as a classification problem and addressed by deep learning. As the convolutional neural network (CNN) has been extensively used in computer vision (CV) and optical metrology, [19][20][21][22] it is also used to learn the mapping between the input wrapped phase and fringe order in phase unwrapping in these methods.…”

Section: Introductionmentioning

confidence: 99%

Hformer: Hybrid convolutional neural network transformer network for fringe order prediction in phase unwrapping of fringe projection

et al. 2022

View full text Add to dashboard Cite

Deep learning based on convolutional neural network (CNN) has attracted more and more attention in phase unwrapping of fringe projection three-dimensional (3D) measurement. However, due to the inherent limitations of convolutional operator, it is difficult to accurately determine the fringe order in wrapped phase patterns that rely on continuity and globality.To attack this problem, in this paper we develop a hybrid CNN-transformer model (Hformer) dedicated to phase unwrapping via fringe order prediction. The proposed Hformer model has a hybrid CNN-transformer architecture that is mainly composed of backbone, encoder, and decoder to take advantage of both CNN and transformer. Backbone is used as a wrapped phase pattern feature extractor. Encoder and decoder with cross attention are designed to enhance global dependency for the fringe order prediction. Experimental results show that the proposed Hformer model achieves better performance in fringe order prediction compared with the CNN models such as U-Net and DCNN. Our work opens an alternative way to the CNN-dominated deep learning phase unwrapping of fringe projection 3D measurement.

show abstract

“…In recent years, deep learning (DL) has been successfully applied in FPP [10]. With deep learning, the absolute phase can be retrieved from two-frequency phase maps [11] or one single-frequency phase map with begin and end fringe order [12]. However, most of the works previously reported on DL-based phase unwrapping in FPP need to acquire large amounts of labeled data to train the models, which, even if available, is laborious and requires professional experts.…”

mentioning

confidence: 99%

Self-supervised phase unwrapping in fringe projection profilometry

Song¹,

Yu²,

TAN³

et al. 2023

Preprint

View full text Add to dashboard Cite

Fast-speed and high-accuracy three-dimensional (3D) shape measurement has been the goal all along in fringe projection profilometry (FPP). The dual-frequency phase shifting (DF-PS) is one of the prominent technologies to achieve this goal. However, the period number of the high-frequency pattern of existing DF-PS approaches is usually limited by the amplified phase errors, setting a limit to measurement accuracy. Deep-learning-based phase unwrapping methods for single-camera FPP usually require labeled data for training. In this letter, a novel self-supervised phase unwrapping method for single-camera FPP systems is proposed. The trained network can retrieve the absolute fringe order from one phase map of 64-period and overperform DF-PS approaches in terms of depth accuracy. Experiments on real scenes demonstrate that the proposed method can unwrap the phase map of motion blur, isolated objects, low reflectivity, and phase discontinuity.

show abstract

Deep absolute phase recovery from single-frequency phase map for handheld 3D measurement

Cited by 22 publications

References 51 publications

PEENet for phase unwrapping in fringe projection profilometry

PEENet for phase unwrapping in fringe projection profilometry

Hformer: Hybrid convolutional neural network transformer network for fringe order prediction in phase unwrapping of fringe projection

Self-supervised phase unwrapping in fringe projection profilometry

Contact Info

Product

Resources

About