针对传统船舶制造检测方法难以满足船体分段复杂曲面高精度检测需求的问题，本文以激光雷达与结构光融合采集的点云为数据源，提出基于曲面曲率的自适应重采样算法，优化点云特征提取流程，并以某船体分段为对象开展对比实验，将本文算法与传统点云处理技术开展对比实验，结果表明，本文方法配准均方根误差小于0.05 mm，偏差检测精度达0.03 mm以内，关键点位一致性系数大于0.95，配准精度、处理效率及超差识别准确率较传统方法均有显著提升，可有效满足船体分段高精度、高效率检测需求。
To address the problem that traditional inspection methods in ship manufacturing cannot meet the high-precision inspection requirements for complex curved surfaces of hull segments, point clouds acquired by the fusion of LiDAR and structured light scanning are used as the data source in this study. An adaptive resampling algorithm based on surface curvature is proposed, and the point cloud feature extraction process is optimized. Comparative experiments are conducted on a real hull segment, in which the proposed algorithm is compared with conventional point cloud processing techniques. The results demonstrate that the registration root-mean-square error of the proposed method is better than 0.05 mm, the deviation detection accuracy is within 0.03 mm, and the consistency coefficient of key points is higher than 0.95. The registration accuracy, processing efficiency, and out-of-tolerance recognition accuracy are all significantly improved compared with the traditional method, which can effectively satisfy the high-precision and high-efficiency inspection requirements of hull segments.
2026,48(8): 180-184 收稿日期：2025-9-4
DOI：10.3404/j.issn.1672-7649.2026.08.028
分类号：U671.99
基金项目：中国高校产学研创基金智能物联网创新教育专项（2024WA078）；武汉船舶职业技术学院科研项目（2025Z07）
作者简介：黄鋆(1988-),女,硕士,副教授,研究方向为虚拟现实技术及船舶数字媒体技术
参考文献：
[1] 冯晓媛, 石琼. 三维虚拟现实物体表面图像点云数据提取仿真[J]. 计算机仿真, 2024, 41(10): 148-152 FENG X Y, SHI Q. Simulation of point cloud data extraction from 3D virtual reality object surface images[J]. Computer Simulation, 2024, 41(10): 148-152
[2] 曾令权. 基于点云数据的船舶轮廓线测绘研究[J]. 舰船科学技术, 2021, 43(12): 208-210 ZENG L Q. Research on ship contour mapping based on point cloud data[J]. Ship Science and Technology, 2021, 43(12): 208-210
[3] 李燕, 王双亭. 基于多传感器数据融合的船舶三维精密测量技术研究[J]. 舰船科学技术, 2025, 47(23): 179-183 LI Y, WANG S T. Research on ship 3D precision measurement technology based on multi-sensor data fusion[J]. Ship Science and Technology, 2025, 47(23): 179-183
[4] 王朋, 刘梦玫, 安振亮, 等. 基于点云数据空间分析的船舶建造精度动态管理决策支持系统[J]. 互联网周刊, 2025(17): 19-21 WANG P, LIU M M, AN Z L, et al. Decision support system for dynamic management of ship construction accuracy based on point cloud data spatial analysis[J]. China Internet Weekly, 2025(17): 19-21
[5] 孙鸿峰. 船舱内部构件识别与点云重建方法研究[D]. 哈尔滨: 哈尔滨工业大学, 2025.
[6] 顾世民, 刘金锋, 钱天龙, 等. 基于几何特征的船舶大尺度构件快速建模及装配干涉检测方法[J/OL]. 中国机械工程, 2024: 1–17. GU S M, LIU J F, QIAN T L, et al. Rapid modeling and assembly interference detection method for large-scale ship components based on geometric features[J/OL]. China Mechanical Engineering, 2024: 1–17.
