基于多阶段深度学习的反射物体三维表面结构重建

IF 1.1 4区 物理与天体物理 Q4 OPTICS
Wenguo Li, Yuyang Yan, Hongjun Lin, Zeqian Feng
{"title":"基于多阶段深度学习的反射物体三维表面结构重建","authors":"Wenguo Li, Yuyang Yan, Hongjun Lin, Zeqian Feng","doi":"10.1007/s10043-024-00940-1","DOIUrl":null,"url":null,"abstract":"<p>With the development of deep learning and structured light streak projection techniques in three-dimensional (3D) imaging, research on the direct reconstruction of 3D shapes from single-streak images has attracted much attention. However, accurately reconstructing 3D shapes is particularly challenging when dealing with objects with specular reflective surfaces. To address this problem, this paper proposes an innovative multi-stage deep learning method that combines the pix2pix adversarial network and a modified version of the DC-HNet architecture. The technique aims to accurately reconstruct 3D shapes from streaked images by eliminating highlights in specular reflection regions through a staged process first. Specifically, the pix2pix adversarial network is first used to eliminate highlights and generate streak images without specular reflections; subsequently, the improved DC-HNet network is further processed to accurately deduce the phase distribution information of the object from the streak images with the elimination of highlights, and then reconstruct the 3D shape. Compared with the traditional self-encoder-based convolutional neural network (CNN) model, the multi-stage approach proposed in this paper significantly improves the accuracy of 3D shape reconstruction by separating the two key steps of highlight elimination and phase derivation and combining them with multi-scale feature enhancement. In this paper, the method’s effectiveness is verified on experimental data, and the results show that the proposed method provides a significant improvement in 3D shape prediction accuracy compared with the existing U-Net network and MultiResUet network. These findings not only demonstrate the innovation and robustness of the proposed method but also show its potential in scientific research and engineering applications.</p>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"80 1 1","pages":""},"PeriodicalIF":1.1000,"publicationDate":"2025-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Three-dimensional surface structure reconstruction of reflective objects using multi-stage deep learning\",\"authors\":\"Wenguo Li, Yuyang Yan, Hongjun Lin, Zeqian Feng\",\"doi\":\"10.1007/s10043-024-00940-1\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>With the development of deep learning and structured light streak projection techniques in three-dimensional (3D) imaging, research on the direct reconstruction of 3D shapes from single-streak images has attracted much attention. However, accurately reconstructing 3D shapes is particularly challenging when dealing with objects with specular reflective surfaces. To address this problem, this paper proposes an innovative multi-stage deep learning method that combines the pix2pix adversarial network and a modified version of the DC-HNet architecture. The technique aims to accurately reconstruct 3D shapes from streaked images by eliminating highlights in specular reflection regions through a staged process first. Specifically, the pix2pix adversarial network is first used to eliminate highlights and generate streak images without specular reflections; subsequently, the improved DC-HNet network is further processed to accurately deduce the phase distribution information of the object from the streak images with the elimination of highlights, and then reconstruct the 3D shape. Compared with the traditional self-encoder-based convolutional neural network (CNN) model, the multi-stage approach proposed in this paper significantly improves the accuracy of 3D shape reconstruction by separating the two key steps of highlight elimination and phase derivation and combining them with multi-scale feature enhancement. In this paper, the method’s effectiveness is verified on experimental data, and the results show that the proposed method provides a significant improvement in 3D shape prediction accuracy compared with the existing U-Net network and MultiResUet network. These findings not only demonstrate the innovation and robustness of the proposed method but also show its potential in scientific research and engineering applications.</p>\",\"PeriodicalId\":722,\"journal\":{\"name\":\"Optical Review\",\"volume\":\"80 1 1\",\"pages\":\"\"},\"PeriodicalIF\":1.1000,\"publicationDate\":\"2025-02-16\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Optical Review\",\"FirstCategoryId\":\"101\",\"ListUrlMain\":\"https://doi.org/10.1007/s10043-024-00940-1\",\"RegionNum\":4,\"RegionCategory\":\"物理与天体物理\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"OPTICS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Optical Review","FirstCategoryId":"101","ListUrlMain":"https://doi.org/10.1007/s10043-024-00940-1","RegionNum":4,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"OPTICS","Score":null,"Total":0}
引用次数: 0

摘要

随着三维成像中深度学习和结构光条投影技术的发展,单条纹图像直接重建三维形状的研究受到了广泛关注。然而,在处理具有镜面反射表面的物体时,准确地重建3D形状尤其具有挑战性。为了解决这个问题,本文提出了一种创新的多阶段深度学习方法,该方法结合了pix2pix对抗网络和DC-HNet架构的改进版本。该技术旨在通过分阶段处理,首先消除镜面反射区域的高光,从条纹图像中精确重建3D形状。具体来说,pix2pix对抗网络首先用于消除高光并生成无镜面反射的条纹图像;随后,对改进的DC-HNet网络进行进一步处理,从剔除高光的条纹图像中准确推断出目标的相位分布信息,重建出三维形状。与传统的基于自编码器的卷积神经网络(CNN)模型相比,本文提出的多阶段方法通过分离高光消除和相位推导两个关键步骤并结合多尺度特征增强,显著提高了三维形状重建的精度。在实验数据上验证了该方法的有效性,结果表明,与现有的U-Net网络和MultiResUet网络相比,该方法在三维形状预测精度上有显著提高。这些发现不仅证明了该方法的创新性和鲁棒性,而且显示了其在科学研究和工程应用方面的潜力。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Three-dimensional surface structure reconstruction of reflective objects using multi-stage deep learning

With the development of deep learning and structured light streak projection techniques in three-dimensional (3D) imaging, research on the direct reconstruction of 3D shapes from single-streak images has attracted much attention. However, accurately reconstructing 3D shapes is particularly challenging when dealing with objects with specular reflective surfaces. To address this problem, this paper proposes an innovative multi-stage deep learning method that combines the pix2pix adversarial network and a modified version of the DC-HNet architecture. The technique aims to accurately reconstruct 3D shapes from streaked images by eliminating highlights in specular reflection regions through a staged process first. Specifically, the pix2pix adversarial network is first used to eliminate highlights and generate streak images without specular reflections; subsequently, the improved DC-HNet network is further processed to accurately deduce the phase distribution information of the object from the streak images with the elimination of highlights, and then reconstruct the 3D shape. Compared with the traditional self-encoder-based convolutional neural network (CNN) model, the multi-stage approach proposed in this paper significantly improves the accuracy of 3D shape reconstruction by separating the two key steps of highlight elimination and phase derivation and combining them with multi-scale feature enhancement. In this paper, the method’s effectiveness is verified on experimental data, and the results show that the proposed method provides a significant improvement in 3D shape prediction accuracy compared with the existing U-Net network and MultiResUet network. These findings not only demonstrate the innovation and robustness of the proposed method but also show its potential in scientific research and engineering applications.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Optical Review
Optical Review 物理-光学
CiteScore
2.30
自引率
0.00%
发文量
62
审稿时长
2 months
期刊介绍: Optical Review is an international journal published by the Optical Society of Japan. The scope of the journal is: General and physical optics; Quantum optics and spectroscopy; Information optics; Photonics and optoelectronics; Biomedical photonics and biological optics; Lasers; Nonlinear optics; Optical systems and technologies; Optical materials and manufacturing technologies; Vision; Infrared and short wavelength optics; Cross-disciplinary areas such as environmental, energy, food, agriculture and space technologies; Other optical methods and applications.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信