An on-line detection technology based on machine vision for micro parts assembly

Abstract

Target preparation technology is one of the core technologies in ICF research. There are many micro parts on the target, the gap between parts is only a few microns, so it is difficult to assemble and realize automatic detection in the assembly process. In this paper, a three-way orthogonal micro vision system is proposed to detect the real-time position and posture of the through-hole and fill-tube of the parts on the target capsule, so to realize the real-time detection in the assembly process. The optical design of the hollow 45-degree reflector is proposed to solve the contradiction between the positive imaging of the through-hole and the positive insertion of the fill-tube. The scheme design of the detection system and the calculation of the system parameters are completed. The precise calibration of the three-dimensional micro vision system and the unified method of the coordinate relationship are proposed. Finally, the experimental test of the online detection system of the target assembly is carried out. The test results show that the system has the following functions. The accuracy of alignment is better than 1μm, and the accuracy of angle alignment is better than 0.2°. It realizes the automation and high-precision detection of the assembly process of small parts, which provides a reliable basis for the automatic assembly of targets for ICF experiments.