Automatic assembly of in-cabin components based on hybrid guidance of binocular vision and laser tracker

Abstract

Abstract In order to automate the assembly of a certain type of missile cabin interior components, a new assembly method with a mixture of binocular vision and laser tracker is proposed. The laser tracker is used to determine the coarse position and posture relationship between the cabin segment and the robot. The two binocular vision systems use skeleton extraction and hit-and-miss transformation to extract target cross-hatch features, perform pose fitting, and obtain the fine pose relationship between the target and the component. Twenty full-process simulations are performed on the CoppeliaSim platform. The results show that, with a pixel accuracy of 0.053 mm, the mean value of the position error of the assembly is 0.614 mm with a standard deviation of 0.270, and the mean value of the posture angle error of the assembly is -0.036° with a standard deviation of 0.178. This method realizes automatic grasp-release action and improved absolute positioning accuracy of the robot in the uncalibrated case and can meet the automation and precision needs of production in the final assembly plant.