影像与激光点云的配准是场景三维重建的关键技术，为爆炸事故溯源分析提供重要的科学支持，并在自动驾驶、灾害溯源等领域展现出广阔的应用前景。然而异源传感器采集的二、三维数据之间存在空间尺度和几何特征的差异，为影像与点云两种模态数据的精细配准带来了困难。鉴于此，本文提出一种基于线特征约束的多视角联合平差配准方法。首先从影像和点云数据中提取线特征并进行粗配准，然后在标准PnP（perspective-n-point）模型基础上引入线特征的方向一致性与正交一致性约束，以最小化多个视角下的误差函数为目标，将变换参数求解转化为非线性最小二乘问题进行迭代优化，最终实现影像与点云的准确配准。该过程无需进行二维-三维投影转换或尺度变换，从而可防止引入投影误差。对比实验结果表明，本文提出的具有线特征约束的多视角联合平差方法可显著提升影像和点云两种模态数据的配准精度。相较于单视角标准PnP模型，本文方法的配准误差降低了60%以上。
The registration of images and laser point clouds is a key technology for 3D scene reconstruction, providing critical scientific support for origin analysis of explosion accidents and holding broad application prospects in fields such as autonomous driving and disaster origin tracing. However, the spatial scale and geometric feature differences between 2D and 3D data collected by heterogeneous sensors pose challenges to the refined registration of image and point cloud modalities. To this end, this paper proposed a multi-view joint adjustment registration method based on line feature constraints. Firstly, line features were extracted from both images and point cloud data for coarse registration. Then, based on the standard perspective-n-point (PnP) model, constraints of directional consistency and orthogonal consistency of line features were introduced. By aiming to minimize the error function across multiple views, the transformation parameter solution was transformed into a nonlinear least squares problem for iterative optimization, ultimately achieving the accurate registration of images and point clouds. This process does not require 2D-3D projection transformations or scale conversions, thus preventing the introduction of projection errors. Comparative experiments show that the proposed multi-view joint adjustment method with line feature constraints can significantly improve the registration accuracy of image and point cloud modalities, reducing registration errors by over 60% compared to the single-view standard PnP model.
[1] 储光涵, 范大昭, 董杨, 等. 结合图论的异源影像点云配准方法[J]. 光学学报, 2023, 43(12): 1228006. Chu G H, Fan D Z, Dong Y, et al. A cross-source image point cloud registration method combined with graph theory [J]. Acta Optica Sinica, 2023, 43(12): 1228006. (in Chinese)
[2] Pujol-Miro A, Ruiz-Hidalgo J, Casas J R. Registration of images to unorganized 3D point clouds using contour cues [C]//25th European Signal Processing Conference (EUSIPCO), 2017: 81-85.
[3] Baghani A, Zoej M J V, Mokhtarzade M. Automatic hierarchical registration of aerial and terrestrial image-based point clouds [J]. European Journal of Remote Sensing, 2018, 51(1): 436-456.
[4] 谢洪, 陈立波, 聂倩, 等. 利用点云配准的空地影像融合技术[J]. 测绘通报, 2022(6): 82-87. Xie H, Chen L B, Nie Q, et al. Air-ground image fusion technology with point cloud registration [J]. Bulletin of Surveying and Mapping, 2022(6): 82-87. (in Chinese)
[5] 李明, 范大昭, 储光涵, 等. 基于相位信息的点云与航空影像配准方法[J]. 测绘地理信息, 2024, 49(3): 74-79. Li M, Fan D Z, Chu G H, et al. A point cloud and aerial image registration method based on phase information [J]. Journal of Geomatics, 2024, 49(3): 74-79. (in Chinese)
[6] 张永军, 洪玮辰, 万一. 利用距离变换模型进行卫星影像与激光点云精配准[J]. 武汉大学学报(信息科学版), 2023, 48(3): 339-348. Zhang Y J, Hong W C, Wan Y. Registration of HRSI and LiDAR point clouds based on distance transformation model [J]. Geomatics and Information Science of Wuhan University, 2023, 48(3): 339-348. (in Chinese)
[7] 胡春梅, 夏国芳, 刘喜. 地面激光点云与数码影像自动高精度配准[J]. 测绘通报, 2021(10): 83-87. Hu C M, Xia G F, Liu X. Automatic high precision registration of terrestrial LiDAR point cloud and digital image [J]. Bulletin of Surveying and Mapping, 2021(10): 83-87. (in Chinese)
[8] 范生宏, 王强, 勾志阳, 等. 一种简易的激光雷达点云与光学影像自动配准方法[J]. 激光杂志, 2021, 42(3): 157-162. Fan S H, Wang Q, Gou Z Y, et al. A simple automatic registration method for lidar point cloud and optical image [J]. Laser Journal, 2021, 42(3): 157-162. (in Chinese)
[9] Crombez N, Caron G, Mouaddib E. 3D point cloud model colorization by dense registration of digital images [C]//Conference on 3D Virtual Reconstruction and Visualization of Complex Architectures (3D-ARCH 2015), 2015: 123-130.
[10] Bennis A, Bombardier V, Thiriet P, et al. Contours based approach for thermal image and terrestrial point cloud registration [C]//24th International CIPA Symposium, 2013: 97-101.
[11] Park J, Kim P, Cho Y K, et al. Framework for automated registration of UAV and UGV point clouds using local features in images [J]. Automation in Construction, 2019, 98: 175-182.
[12] Jeon Y, Seo S W. EFGHNet: a versatile image-to-point cloud registration network for extreme outdoor environment [J]. IEEE Robotics and Automation Letters, 2022, 7(3): 7511-7517.
[13] 赵夫群, 黄鹤, 耿国华. 基于主成分特征向量的点云配准方法[J]. 应用科学学报, 2024, 42(6): 962-976. Zhao F Q, Huang H, Geng G H. Point cloud registration method based on principal component eigenvectors [J]. Journal of Applied Sciences, 2024, 42(6): 962-976. (in Chinese)
[14] Yuan C, Liu X, Hong X, et al. Pixel-level extrinsic self-calibration of high resolution LIDAR and camera in targetless environments [J]. IEEE Robotics and Automation Letters, 2021, 6(4): 7517-7524.
[15] Lepetit V, Moreno-Noguer F, Fua P. EPnP: an accurate O(n) solution to the PnP problem [J]. International Journal of Computer Vision, 2009, 81(2): 155-166.
