桥梁实景三维高斯辐射场建模

doi:10.11947/j.AGCS.2024.20240071

测绘学报 ›› 2024, Vol. 53 ›› Issue (9): 1694-1705.doi: 10.11947/j.AGCS.2024.20240071

• 精密工程测量 • 上一篇

桥梁实景三维高斯辐射场建模

马威¹^,²^,³(), 涂强¹, 潘建平¹^,²^,³(), 赵立都¹, 涂伟⁴^,⁵^,⁶^,⁷, 李清泉⁴^,⁵^,⁶^,⁷

^1.重庆交通大学智慧城市学院，重庆　400074
^2.自然资源部国土空间规划监测评估预警重点实验室，重庆　400074
^3.自然资源部智能城市时空信息与装备工程技术创新中心，重庆　401121
^4.深圳大学建筑与城市规划学院城市空间信息工程系，广东　深圳　518060
^5.空间信息智能感知与服务深圳市重点实验室，广东　深圳　518060
^6.广东省城市空间信息工程重点实验室，广东　深圳　518060
^7.自然资源部大湾区地理环境监测重点实验室，广东　深圳　518060

收稿日期:2024-02-21 发布日期:2024-10-16
通讯作者: 潘建平 E-mail:weima@cqjtu.edu.cn;panjianping@qq.com;panJianping@qq.com
作者简介:马威（1987—），男，博士，副教授，研究方向为智能测绘与场景感知、视觉三维重建、地理增强现实。E-mail：weima@cqjtu.edu.cn
基金资助:
自然资源部国土空间规划监测评估预警重点实验室开放基金资助项目(LMEE-KF2023004);自然资源部智能城市时空信息与装备工程技术创新中心开放基金(STIEIC-KF202305);国家自然科学青年基金(42001324);重庆市教育委员会科学技术研究项目(KJQN202200744);重庆市自然科学基金(cstc2021jcyj-msxmX1147);宁夏回族自治区重点研发计划(2022CMG02014);重庆市研究生联合培养基地建设项目(JDLHPYJD2019004)

3D Gaussian radiation field modeling for real-scene bridges

Wei MA¹^,²^,³(), Qiang TU¹, Jianping PAN¹^,²^,³(), Lidu ZHAO¹, Wei TU⁴^,⁵^,⁶^,⁷, Qingquan LI⁴^,⁵^,⁶^,⁷

^1.College of Smart City, Chongqing Jiaotong University, Chongqing 400074, China
^2.Key Laboratory of Land Space Planning Monitoring, Evaluation, and Early Warning, Ministry of Natural Resources, Chongqing 400074, China
^3.Engineering Technology Innovation Center for Smart City Spatiotemporal Information and Equipment, Ministry of Natural Resources, Chongqing 401121, China
^4.Department of Urban Spatial Information Engineering, School of Architecture and Urban Planning, Shenzhen University, Shenzhen 518060, China
^5.Key Laboratory of Intelligent Perception and Services for Spatial Information, Shenzhen 518060, China
^6.Key Laboratory of Urban Spatial Information Engineering of Guangdong Province, Shenzhen 518060, China
^7.Key Laboratory of Geographic Environment Monitoring in the Greater Bay Area, Ministry of Natural Resources, Shenzhen 518060, China

Received:2024-02-21 Published:2024-10-16
Contact: Jianping PAN E-mail:weima@cqjtu.edu.cn;panjianping@qq.com;panJianping@qq.com
About author:MA Wei (1987—), male, PhD, associate professor, majors in intelligent remote sensing methods and applications, spatiotemporal big data engineering. E-mail: weima@cqjtu.edu.cn
Supported by:
The Open Fund of the Key Laboratory of Monitoring, Evaluation, and Early Warning of Territorial Spatial Planning, Ministry of Natural Resources(LMEE-KF2023004);The Open Fund of the Engineering Technology Innovation Center for Smart City Spatio-temporal Information and Equipment, Ministry of Natural Resources(STIEIC-KF202305);The National Natural Science Foundation of China(42001324);The Science and Technology Research Project of Chongqing Education Commission(KJQN202200744);Chongqing Natural Science Foundation(cstc2021jcyj-msxmX1147);The Key Research and Development Program of Ningxia Hui Autonomous Region(2022CMG02014);Chongqing Graduate Joint Training Base Construction Project(JDLHPYJD2019004)

摘要/Abstract

摘要：

实景三维和数字孪生已成为桥梁运维和管理的重要基础，然而面对桥梁的复杂几何结构，现有三维建模方法存在原始数据采集量大、建模效率低、模型细节缺失或变形等问题。对此，本文研究了一种基于3D高斯辐射场的桥梁实景三维重建方法。利用3D高斯函数对采集图像生成的稀疏点云构建高斯辐射场，并基于随机梯度下降对辐射场参数进行自适应优化，通过可微光栅渲染对三维模型进行实时可视化，从而实现高质量桥梁三维重建和渲染。试验探讨了不同图像分辨率，以及各项参数变化对桥梁建模的影响，并与传统方法作对比，为进一步针对桥梁的应用提供理论和技术支撑，推动桥梁复杂结构的高效、准确实景三维重建。

关键词: 3D高斯辐射场, 桥梁实景三维重建, 随机梯度下降, 可微光栅渲染

Abstract:

Realistic 3D modeling and digital twins have become essential foundations for bridge operation and management. However, given the complex geometric structures of bridges, current 3D modeling methods face issues such as large amounts of raw data collection, low modeling efficiency, and missing or deformed model details. In response to these challenges, this paper investigates a bridge realistic 3D reconstruction method based on 3D Gaussian radiance fields. This method utilizes 3D Gaussian functions to construct a Gaussian radiance field from sparse point clouds generated by captured images. Adaptive optimization of radiance field parameters is performed based on stochastic gradient descent, and real-time visualization of the 3D model is achieved through differentiable rasterization, resulting in high-quality bridge 3D reconstruction and rendering. The study explores the impact of different image resolutions and various parameter changes on bridge modeling. Comparisons with traditional methods are made to provide theoretical and technical support for further bridge applications, promoting efficient and accurate realistic 3D reconstruction of complex bridge structures.

Key words: 3D Gaussian splatting, 3D bridge reconstruction, stochastic gradient descent, differentiable rasterizer rendering

中图分类号:

P258

马威, 涂强, 潘建平, 赵立都, 涂伟, 李清泉. 桥梁实景三维高斯辐射场建模[J]. 测绘学报, 2024, 53(9): 1694-1705.

Wei MA, Qiang TU, Jianping PAN, Lidu ZHAO, Wei TU, Qingquan LI. 3D Gaussian radiation field modeling for real-scene bridges[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(9): 1694-1705.

图/表 13

图1

图2

图3

图4

表1

图5

图6

图7

图8

图9

图10

表2

图11

参考文献 32

[1]	涂伟, 李清泉, 高文武, 等. 基于机器视觉的桥梁挠度实时精密测量方法[J]. 测绘地理信息, 2020, 45(6):80-87.
	TU Wei, LI Qingquan, GAO Wenwu, et al. Monitoring the dynamic deflection of bridges using computer vision[J]. Journal of Geomatics, 2020, 45(6):80-87.
[2]	朱庆, 张利国, 丁雨淋, 等. 从实景三维建模到数字孪生建模[J]. 测绘学报, 2022, 51(6):1040-1049. DOI: 10.11947/j.AGCS.2022.20210640.
	ZHU Qing, ZHANG Liguo, DING Yulin, et al. From real 3D modeling to digital twin modeling [J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(6):1040-1049. DOI: 10.11947/j.AGCS.2022.20210640.
[3]	李德仁, 肖雄武, 郭丙轩, 等. 倾斜影像自动空三及其在城市真三维模型重建中的应用[J]. 武汉大学学报(信息科学版), 2016, 41(6):711-721.
	LI Deren, XIAO Xiongwu, GUO Bingxuan, et al. Oblique image based automatic aerotriangulation and its application in 3D city model reconstruction[J]. Geomatics and Information Science of Wuhan University, 2016, 41(6):711-721.
[4]	王云川, 段平, 李佳, 等. 不同航线规划的无人机影像三维建模质量分析[J]. 遥感信息, 2020, 35(2):121-126.
	WANG Yunchuan, DUAN Ping, LI Jia, et al. Analysis on 3D modeling quality of UAV images for different route planning[J]. Remote Sensing Information, 2020, 35(2):121-126.
[5]	GOESELE M, SNAVELY N, CURLESS B, et al. Multi-view stereo for community photo collections[C]//Proceedings of 2007 IEEE 11th International Conference on Computer Vision. Rio de Janeiro: IEEE, 2007: 1-8.
[6]	陈正飞. 基于无人机摄影测量的桥梁实景三维重建[J]. 河南科技, 2023, 58(14):15-20.
	CHEN Zhengfei. 3D reconstruction of bridge real scene based on UAV photogrammetry [J]. Henan Science and Technology, 2023, 58(14):15-20.
[7]	毛琳, 王超, 余星宇, 等. 基于无人机倾斜摄影的桥梁三维建模研究[J]. 港工技术, 2022, 59(6):58-62.
	MAO Lin, WANG Chao, YU Xingyu, et al. Research on 3D modeling of bridge based on UAV tilt photography[J]. Port Engineering Technology, 2022, 59(6):58-62.
[8]	崔庆宏. 精细化建模在特大桥维护检测中的应用[J]. 测绘通报, 2021(5):162-166.
	CUI Qinghong. Application of refined modelling in maintenance and detection for extra large bridge[J]. Bulletin of Surveying and Mapping, 2021(5):162-166.
[9]	俞建康. 无人机倾斜摄影测量在立交桥三维重建中的应用[J]. 测绘通报, 2020(10):106-109.
	YU Jiankang. Application of UAV tilt photogrammetry in 3D reconstruction of overpass[J]. Bulletin of Surveying and Mapping, 2020(10):106-109.
[10]	李清泉, 黄惠, 姜三, 等. 优视摄影测量方法及精度分析[J]. 测绘学报, 2022, 51(6):996-1007. DOI: 10.11947/j.AGCS.2022.20210685.
	LI Qingquan, HUANG Hui, JIANG San, et al. Optimized views photogrammetry and its precision analysis[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(6):996-1007. DOI: 10.11947/j.AGCS.2022.20210685.
[11]	肖雄武. 具备结构感知功能的倾斜摄影测量场景三维重建[J]. 测绘学报, 2019, 48(6):802. DOI: 10.11947/j.AGCS.2019.20180319.
	XIAO Xiongwu. Oblique photogrammetry based scene 3D reconstruction with structure sensing functions[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(6):802. DOI: 10.11947/j.AGCS.2019.20180319.
[12]	ZHAO Sizeng, KANG Fei, LI Junjie, et al. Structural health monitoring and inspection of dams based on UAV photogrammetry with image 3D reconstruction[J]. Automation in Construction, 2021, 130:103832.
[13]	SNAVELY N, SEITZ S M, SZELISKI R. Photo tourism: exploring photo collections in 3D[M]//Seminal Graphics Papers: pushing the Boundaries, Volume 2. New York: ACM Press, 2023: 515-526.
[14]	WANG Feng, ZOU Yang, DEL REY CASTILLO E, et al. Automated UAV path-planning for high-quality photogrammetric 3D bridge reconstruction[J]. Structure and Infrastructure Engineering, 2022:1-20.
[15]	陈卓, 马洪超. 基于机载LiDAR数据的大型立交桥自动提取与建模方法[J]. 测绘学报, 2012, 41(2):252-258.
	CHEN Zhuo, MA Hongchao. Automatic extracting and modeling approach of city cloverleaf from airborne LiDAR data[J]. Acta Geodaetica et Cartographica Sinica, 2012, 41(2):252-258.
[16]	李卉, 钟成, 黄先锋, 等. 集成激光雷达数据和遥感影像的立交桥自动检测方法[J]. 测绘学报, 2012, 41(3):428-433.
	LI Hui, ZHONG Cheng, HUANG Xianfeng, et al. Automatic overpass detection with LiDAR and remote sensing image[J]. Acta Geodaetica et Cartographica Sinica, 2012, 41(3):428-433.
[17]	MEHRANFAR M, AREFI H, ALIDOOST F. Knowledge-based 3D reconstruction of bridge structures using UAV-based photogrammetric point cloud[J]. Journal of Applied Remote Sensing, 2021, 15(4):044503.
[18]	MILDENHALL B, SRINIVASAN P P, TANCIK M, et al. Nerf: representing scenes as neural radiance fields for view synthesis[C]//Proceedings of 2020 European Conference on Computer Vision. Cham: Springer, 2020.
[19]	YU A, YE V, TANCIK M, et al. PixelNeRF: neural radiance fields from one or few images[C]//Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville: IEEE, 2021: 4578-4587.
[20]	CHEN Anpei, XU Zexiang, ZHAO Fuqiang, et al. MVSNeRF: fast generalizable radiance field reconstruction from multi-view stereo[C]//Proceedings of 2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 14124-14133.
[21]	WANG Qianqian, WANG Zhicheng, GENOVA K, et al. IBRNet: learning multi-view image-based rendering[C]//Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville: IEEE, 2021: 4690-4699.
[22]	CHEN Zhiqin, FUNKHOUSER T, HEDMAN P, et al. MobileNeRF: exploiting the polygon rasterization pipeline for efficient neural field rendering on mobile architectures[C]//Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 16569-16578.
[23]	TEWARI A, THIES J, MILDENHALL B, et al. Advances in neural rendering[J]. Computer Graphics Forum, 2022, 41(2):703-735.
[24]	ZHANG Kai, RIEGLER G, SNAVELY N, et al. NeRF++: analyzing and improving neural radiance fields[EB/OL]. [2024-03-25]. https://arxiv.org/abs/2010.07492v2.
[25]	TURKI H, RAMANAN D, SATYANARAYANAN M. Mega-NeRF: scalable construction of large-scale NeRFs for virtual fly- throughs[C]//Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022.
[26]	MÜLLER T, EVANS A, SCHIED C, et al. Instant neural graphics primitives with a multiresolution hash encoding[J]. ACM Transactions on Graphics, 2022, 41(4):1-15.
[27]	KERBL B, KOPANAS G, LEIMKUEHLER T, et al. 3D Gaussian splatting for real-time radiance field rendering[J]. ACM Transactions on Graphics, 2023, 42(4):1-14.
[28]	YANG Ziyi, GAO Xinyu, ZHOU Wen, et al. Deformable 3D Gaussians for high-fidelity monocular dynamic scene reconstruction[EB/OL]. [2024-01-30]. https://arxiv.org/abs/2309.13101v2.
[29]	RHODIN H, ROBERTINI N, RICHARDT C, et al. A versatile scene model with differentiable visibility applied to generative pose estimation[C]//Proceedings of 2015 IEEE International Conference on Computer Vision. Santiago: IEEE, 2015.
[30]	YU Z, CHEN A, HUANG B, et al. Mip-splatting: alias-free 3D Gaussian splatting[C]//Proceedings of 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. SEATTLE: IEEE, 2024.
[31]	CHEN Yu, LEE G H. DoGaussian: distributed-oriented Gaussian splatting for large-scale 3D reconstruction via Gaussian consensus[EB/OL]. [2024-04-13]. https://arxiv.org/abs/2405.13943v1.
[32]	HENZLER P, MITRA N, RITSCHEL T. Escaping Plato's cave: 3D shape from adversarial rendering[C]//Proceedings of 2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019.

λ	PSNR↑	SSIM↑	LPIPS↓
0	27.154	0.815	0.363
0.1	27.775	0.837	0.324
0.2	27.908	0.843	0.310
0.3	27.892	0.841	0.306
0.4	27.671	0.841	0.303
0.5	27.659	0.840	0.300
0.6	27.278	0.835	0.301
0.7	27.118	0.831	0.302
0.8	27.047	0.831	0.301
0.9	26.866	0.829	0.302
1.0	26.248	0.822	0.305

迭代次数	PSNR↑	SSIM↑	LPIPS↓	训练时间	Loss
6000	24.197	0.747	0.459	3 min 8 s	0.079
12 000	25.710	0.794	0.389	7 min 32 s	0.047
18 000	26.338	0.813	0.359	13 min 1 s	0.048
24 000	26.916	0.820	0.347	18 min 28 s	0.039
30 000	27.454	0.824	0.339	23 min 18 s	0.041
36 000	27.349	0.824	0.339	28 min 45 s	0.038

桥梁实景三维高斯辐射场建模

3D Gaussian radiation field modeling for real-scene bridges

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 32

相关文章 15

编辑推荐

Metrics

本文评价

[1]	范千, 方绪华, 许承权, 杨荣华. 变形监测数据预报的动态贝叶斯ELM方法[J]. 测绘学报, 2019, 48(7): 919-925.
[2]	李阳腾龙. 高速铁路轨道精测精调及其平顺性优化研究[J]. 测绘学报, 2018, 47(11): 1562-1562.
[3]	李清泉, 毛庆洲. 道路/轨道动态精密测量进展[J]. 测绘学报, 2017, 46(10): 1734-1741.
[4]	向巍. 测量数据拟合理论与方法及其在构件建模中的应用[J]. 测绘学报, 2017, 46(9): 1203-1203.
[5]	魏冠军, 党亚民, 章传银, 杨维芳. 顾及不确定性影响的变形概率预报法[J]. 测绘学报, 2017, 46(4): 526-532.
[6]	尹晖, 周晓庆, 张晓鸣. 非等间距多点变形预测模型及其应用[J]. 测绘学报, 2016, 45(10): 1140-1147.
[7]	余加勇. 基于GNSS和RTS技术的桥梁结构健康监测[J]. 测绘学报, 2015, 44(10): 1177-1177.
[8]	李珵, 卢小平, 朱宁宁, 卢遥, 武永斌, 李国清. 基于激光点云的隧道断面连续提取与形变分析方法[J]. 测绘学报, 2015, 44(9): 1056-1062.
[9]	黎奇, 白征东, 陈波波, 过静珺, 辛浩浩, 程宇航, 黎琼, 吴斐. GNSS/INS多传感器组合高速铁路轨道测量系统[J]. 测绘学报, 2020, 49(5): 569-579.
[10]	郭迎钢, 李宗春, 何华, 王志颖. 三维坐标转换公共点最优权值的单纯形搜索算法[J]. 测绘学报, 2020, 49(8): 1004-1013.
[11]	郭迎钢, 李宗春, 何华, 张冠宇, 冯其强, 杨浩. 变形监测网稳定点选取的平方型M_split相似变换法[J]. 测绘学报, 2020, 49(11): 1419-1429.
[12]	李清泉, 张德津, 汪驰升, 陈智鹏, 涂伟. 动态精密工程测量技术及应用[J]. 测绘学报, 2021, 50(9): 1147-1158.
[13]	李冠青. 特长沉管隧道高精度贯通测控方法研究[J]. 测绘学报, 2023, 52(5): 869-869.
[14]	朱宁宁, 杨必胜, 董震. 基于遥感的森林碳储量显式计算模型[J]. 测绘学报, 2024, 53(1): 36-49.
[15]	潘雄, 黄伟凯, 赵万卓, 张思莹, 张龙杰, 金丽宏. 基于BiLSTM模型的BDS-3短期钟差预报精度研究[J]. 测绘学报, 2024, 53(1): 65-78.