Research On Position And Installation Of Ship Propulsion Shafting System Based On Vision System

Prior to this,the six-degree-of-freedom actuator used a laser collimator to position and install it through a vision-assisted shafting system.The positioning accuracy was poor.Later,a three-dimensional vision sensor was introduced in the distributed six-degree-of-freedom positioning system of the ship’s shafting system to improve the system’s axis The spatial perception ability of the system pose improves the positioning accuracy of the shaft system,and improves the efficiency of shaft positioning,alignment and installation.In this paper,a three-dimensional laser scanner is applied to the alignment of the shafting system.Through the processing of the three-dimensional point cloud data of the shafting system,the relative position and posture of the two shafting systems are obtained to complete the identification and positioning of the shafting system.The main research work in this paper is as follows Show:First,the three-dimensional point cloud information acquisition of ship shafting based on lateral movement is studied.By analyzing the three-dimensional laser scanner based on the laser triangulation method,a laterally moving three-dimensional laser scanning system is established,the hardware components of the system are introduced,the system coordinate system is established,and the image coordinate system,scanner coordinate system,and actuator coordinates are studied The conversion relationship of the three major coordinate systems.Secondly,research the pre-processing method of image acquisition.Preprocessing is mainly divided into two parts: filtering processing and downsampling processing.Firstly,the external interference points are roughly screened through conditional filtering,and the interference points near the main body of the image are screened based on the distributed distance method based on the principle of using K-D tree to find neighboring points.The downsampling process based on step size distribution method and voxelization grid method is studied to reduce the point cloud density.Again,the boundary point cloud information of the processed point cloud data is extracted.The edge cloud feature extraction method based on scan line data is introduced.The mesh partition method is used to complete the boundary extraction of point cloud data.The boundary point cloud is fitted as a straight line,and the difference between the position and pose of the two-axis point cloud is obtained to complete the point cloud data processing.Finally,it is studied whether the alignment of the axle system after the alignment has reached the accuracy.Extract the data of the point cloud data of the shaft system after alignment,compare the straight line slope and relative position relationship of the point cloud of the flange and shaft system boundary,and complete the quality inspection of the shaft system after alignment.