针对船舱复杂构件点云提取存在人工成本高、效率低的问题,提出了一种适用于平面舱壁类型船舱点云的分割方法。通过种子点集构建、点云法线估计及直线拟合的方式建立以船舱纵向为X轴、横向为Y轴、竖向为Z轴的独立坐标系,以简化分割算法的复杂度;根据船舱内部复杂构件的分布特性,制定最佳分割次序,基于随机采样一致性算法拟合平面的思想有序地分割船舱构件点云。选用两组不同结构的船舱点云数据进行算法验证,实验结果表明:该方法能够从不同结构的船舱散乱点云中快速、准确地自动分割出主要构件点云,可靠性强,具有较高的实用价值。 Aiming at the problem of high manual labor cost and low efficiency of point cloud extraction of complex component of cabin, a new method of segment point cloud suitable for plane bulkhead is proposed. Through the establishment of seed point set, point cloud normal estimation and straight line fitting, an independent coordinate system with longitudinal X axis, horizontal Y axis and vertical Z axis is established to simplify the complexity of the segmentation algorithm. The distribution characteristics of the internal complex components, the best partitioning order was formulated, and the point cloud of the cabin components was orderly partitioned based on the idea of ​​fitting the plane by random sampling consistency algorithm. Two groups of cabin point cloud data with different structures are selected to verify the algorithm. The experimental results show that this method can quickly and accurately automatically separate the point clouds of main components from the scattered point cloud of different structures, which has high reliability and high reliability Practical value.