Bathymetric LiDAR and multibeam echo-sounding data registration methodology employing a point cloud model. (June 2022)

• Record Type: Journal Article
• Authors: Li, Shaoyu; Su, Dianpeng; Yang, Fanlin; Zhang, Huijuan; Wang, Xiankun; Guo, Yadong
• Is Part Of: Applied ocean research. Volume 123(2022)
• Publication Date: 2022-06
• Subjects: Multiplatform point cloud registration -- Airborne LiDAR bathymetry -- Shipborne multibeam sounder -- Topographic features; Ocean engineering -- Periodicals; 620.416205
• DOI: 10.1016/j.apor.2022.103147
• Languages: English
• ISSNs: 0141-1187

Abstract

Highlights: Realize the full coverage and integration mapping of seabed topography by fusing the ALB and MBES point cloud data. The azimuth angle features and curvature feature are used to find the corresponding feature points from the source and target point cloud data. The edge key feature points are extracted to reduce the influence of inconsistent density in multisource data and gaps in the data on registration. The registration accuracy is improved compared to the results obtained with the traditional curvature-based registration (CR) algorithm. Abstract: Airborne LiDAR bathymetry (ALB) and multibeam echosounders (MBESs) are the most commonly used technologies for underwater topographical measurements. However, ALB cannot obtain underwater information in deep water. Due to concerns about the safety of the survey vessels, it is difficult to take measurements in shallow water using an MBES. To maximize the complementary advantages of these two technologies, this paper proposes a point cloud registration algorithm for ALB and MBES data based on azimuth angle features. First, the topographical curvature features of ALB and the MBES point cloud data are extracted by using a cylindrical neighborhood model and a quadric surface fitting algorithm. Then, the azimuth angle similarity of the curvature feature points is used to match the corresponding points. Finally, a truncated least-squares-based iterative nearest-adjacent point (TrICP) method is utilized to realize multisource data registration. To verify the performance of the proposed registration algorithm, four areas with typical ALB and MBES point cloud data from Xisha Island, South China Sea, are used as the study areas. Compared to the traditional and typical curvature-based registration (CR) algorithm for point cloud data registration, the proposed algorithm obtains average RMSE values of 0.278 m, 0.252 m, 0.214 m and 0.177 m in the four typical study areas. The corresponding registration accuracy is improved by 91%, 73%, 84% and 49%, respectively. The experimental results show that the proposed algorithm is more reliable and effective for ALB and MBES point cloud registration than the previous algorithms.