一种优化的船体外板三维点云数据提取方法

针对水火弯板机检测系统中船舶外板三维点云数据自动提取过程存在边缘噪点、板下贴合垫木识别效率低以及外板边缘拟合等问题,提出一种优化DBSCAN聚类算法,首先根据现场加工环境精简点云,利用网格划分来建立外板点云拓扑模型,然后根据DBSCAN密度聚类算法搜索出外板点云,最后采用最小二乘法进行边缘拟合。结果表明,该算法能有效识别外板边缘,提取出外板点云。

An optimization of hull outer steel plate three-dimensional extraction method for point cloud data

For plate bending by line heating machine detection system of ship hull plate in three dimensional automatic recognition and extraction of point cloud data plate edge noise, under the plate laminating wood identifies inefficient or not recognized as well as external plate edge point cloud fitting and so on. Paper proposed has a optimization DBSCAN poly class algorithm, and according to outside board processing environment combines thought, first according to outside board processing site environment streamlining points cloud data, established using network grid divided to established outside board points cloud topology model, then this based Shang according to based on density of poly class method search out belongs to outside board points cloud, findedge, last used minimum II multiplication for edge intends collection. Experimental results show that the algorithm can effectively remove outer boundary point cloud noise, recognition outside board edges, extracting out points.