为保障船舶移动过程中的安全性，提出船舶移动路径三维动态场景重建方法。船舶航行过程中，通过相机等视觉传感器在未知环境下连续获取船舶移动路径图像；采用基于ORB特征提取的算法提取船舶移动路径图像内的特征点，根据汉明距离进行特征点匹配；通过对极几何解算特征点匹配结果获取相机的位姿信息；采用基于特征相关性筛选的关键帧选取机制获取全部关键帧，通过关键帧的点云拼接实现船舶移动路径三维动态场景重建。仿真结果显示，该方法不仅能够有效实现特征点匹配，且正确匹配率达到97%以上，三维场景重建结果的均方误差控制在0.2以下，结构相似性始终高于95%。
To ensure the safety of ship movement, a three-dimensional dynamic scene reconstruction method for ship movement path is studied. During ship navigation, visual sensors such as cameras are used to continuously obtain images of the ship's movement path in unknown environments; Using an ORB based feature extraction algorithm to extract feature points within the ship's movement path image, and matching feature points based on Hamming distance; Obtain camera pose information by calculating feature point matching results through polar geometry; Adopting a keyframe selection mechanism based on feature correlation filtering to obtain all keyframes, the 3D dynamic scene reconstruction of ship movement path is achieved through point cloud concatenation of keyframes. The simulation results show that this method can not only effectively achieve feature point matching, but also achieve a correct matching rate of over 97%. The mean square error of the 3D scene reconstruction results is controlled below 0.2, and the structural similarity is always above 95%.
2024,46(14): 166-169 收稿日期：2024-01-09
DOI：10.3404/j.issn.1672-7649.2024.14.028
分类号：TP18
作者简介：熊媛媛(1981-),女,博士,讲师,研究方向为虚拟现实及计算机三维设计
参考文献：
[1] 孙凌宇, 秦红亮, 师瑞卓, 等. 基于激光雷达的移动剖面式三维形貌重建方法研究[J]. 应用激光, 2023, 43(10): 136-147.
SUN Lingyu, QIN Hongliang, SHI Ruizhuo, et al. Research on moving profile 3D shape reconstruction method based on lidar[J]. Applied Laser, 2023, 43(10): 136-147.
[2] KARAMI A , MENNA F , REMONDINO F , et al. Exploiting light directionality for image-based 3D reconstruction of non-collaborative surfaces[J]. Photogrammetric Record, 2022, 37(5): 111–138.
[3] 李翔宇, 张雪芹. ORBTSDF-SCNet: 一种动态场景在线三维重建方法[J]. 华东理工大学学报(自然科学版), 2023, 49(2): 284-294.
LI Xiangyu, ZHANG Xueqin. ORBTSDF-SCNet: An online 3d reconstruction method for dynamic scene[J]. Journal of East China University of Science and Technology(Natural Science Edition), 2023, 49(2): 284-294.
[4] 江瑞祥, 缪君, 储珺, 等. 基于多尺度聚焦网络的单图像城市场景3D平面重建[J]. 小型微型计算机系统, 2022, 43(8): 1718-1724.
JIANG Ruixiang, MIAO Jun, CHU Jun, et al. 3D planar reconstruction of urban scene from single image based on multi-scale focusing network[J]. Journal of Chinese Computer Systems, 2022, 43(8): 1718-1724.
[5] 许少秋, 余扬帆, 郭俊林, 等. 基于大场景三维点云的集装箱船关键绑扎特征识别与重建[J]. 船舶工程, 2022, 44(6): 127-133.
XU Shaoqiu, YU Yangfan, GUO Junlin, et al. Identification and reconstruction of key lashing features of container ships based on large scene 3D point cloud[J]. Ship Engineering, 2022, 44(6): 127-133.
[6] 王超, 杨龙, 吴雪峰, 等. 场景目标的稀疏序列融合三维扫描重建[J]. 计算机辅助设计与图形学学报, 2023, 35(6): 819-829.
WANG Chao, YANG Long, WU Xuefeng, et al. 3D scanning of scene-level targets based on the sparse sequence fusion[J]. Journal of Computer-Aided Design & Computer Graphics, 2023, 35(6): 819-829.
