| In 3d measurement system,linear structured light vision system has been widely studied and applied for its advantages of non-contact,high controllability,high real-time performance,high precision,wide measurement range and simple structure.Through the analysis of various types of active three dimensional inspection technology,the line structured light threedimensional measurement system has the characteristics of simple structure,fast scanning and data processing speed,high precision,relatively low equipment cost,and has been widely used in product quality control and reverse engineering.Its basic idea is to project the laser beam onto the surface of the object to be measured,and map a bright laser line on the surface of the object to be measured.From different projection directions,the laser line will produce distortion due to the change of the height of the measured object surface.By calculating the coordinates of the bright line image on the image plane,the depth data of a section of the object can be solved,and then by introducing one-dimensional moving scan,the three-dimensional shape distribution of the object can be obtained.Based on the research and analysis of structural light vision,simulation technology and robot application at home and abroad,this paper designs a set of visual scheme of robot automatic production line suitable for polishing small ship components and carries out experimental research.The main research contents and results are as follows:First of all,combined with the measurement requirements of the system,by comparing various types of 3D visual measurement methods and their advantages and disadvantages,on this basis,it is analyzed that the line structured light visual measurement method is more suitable for the ship small component grinding production line vision scheme,and then carries out the overall scheme design of the system,formulate the working process of the system.The system is divided into three parts:image acquisition module,data processing module and actuator module.According to the design scheme,the system hardware selection,including industrial camera,lens,laser,robot and other parts,and the hardware platform was built,as well as a detailed description of the system software module functions,QT software platform to complete the design of software module and human-computer interaction interface design.Then,by studying the measurement algorithm of line structured light vision,the measurement model of line structured light vision is established and analyzed to complete the calibration of system parameters.The processing method of light strip image is analyzed,and the research of image denoising,image enhancement,threshold segmentation and edge detection is completed.In laser strip center detection,on the basis of summarizing all kinds of strip center extraction algorithms and their advantages and disadvantages,an improved gray center method is proposed to extract the center of the strip.Finally,through the analysis and extraction of point cloud data,the 3d reconstruction of the segmented workpiece surface contour is carried out.Finally,a hardware-in-loop simulation platform of ship small component grinding based on line structured light vision system was built for the application of the system,including a virtual workstation based on Robotstudio and ABB IRB1200 robot physical workstation.The function of each part of the system was tested by taking the hull circular sheet as the experimental object.The feasibility of the system is verified by the experimental process including system parameter calibration,system communication and robot execution.Based on the design and implementation process of the integrated system and the final test results,the vision system based on line structured light studied in this paper is of great significance to improve the grinding efficiency and product quality of production line,and provides valuable development direction and application cases for the visual scheme application of robotic automatic production line for the grinding of small Marine components. |