Pipeline Weld And Recognition Based On Stereo Vision And Laser Ranging

As the new generation of satellite platform has higher requirements for the production cycle and welding quality of the propulsion system pipeline,and the requirements for the automation degree of the welding equipment continue to improve,the traditional satellite propulsion system pipeline welding is completed by manual operation,which cannot meet the requirements in terms of quality,production cycle and reliability,and the accuracy depends on the experience of the operator.In order to solve the above problems,it is necessary to develop an automatic welding platform based on binocular stereo vision,which can automatically locate the weld,automatically fix and automatically complete the tungsten pole alignment,so as to realize the automatic welding of a large number of welds in the satellite propulsion system.Under the background of this practical project,this paper finally completed the design of the algorithm of weld position coordinate recognition based on binocular stereo vision and laser ranging.The research of this project is of great significance to improve the production efficiency of welding work.In this paper,the current situation of binocular stereo vision and welding robot at home and abroad is analyzed.According to the overall design index and design principle of the system,the hardware selection and installation of the related hardware are carried out.On this basis,the simulation environment of the vision system is built.The transformation relationship between camera imaging coordinate system and the establishment of imaging model are studied,as well as the imaging model of binocular stereo vision system.At last,the algorithm principle of monocular camera calibration,binocular camera calibration and calibration,robot hand eye calibration are studied,and all the calibration and calibration work of the system is completed.On the basis of image preprocessing,the algorithm to determine the region of titanium alloy conduit is studied and designed.In this paper,the algorithm principle of feature extraction and stereo matching is studied,and the SIFT algorithm,surf algorithm,orb algorithm and RANSAC algorithm are combined to test the feasibility of the algorithm.In addition,two kinds of attitude of the conduit arecalculated,including coplanar rotation and plane rotation.After the rotation,the algorithm of calculating the coordinates of the conduit weld is studied and designed.Finally,the feasibility and stability of the algorithm are verified by experiments.At last,all the modules of the system are integrated,and the software of vision system and manipulator control system is designed and compiled.After the design,the relevant experiments are carried out,and the functions of binocular camera image acquisition,welding clamp motor control,manipulator drive control,laser distance measurement,conduit attitude estimation and weld position recognition are completed.Through the accuracy analysis of the whole system It is known that the system has good stability and feasibility.