The Design and Research of a Cup Body Weld Seam Polishing Positioning System Based on Machine Vision

Abstract

A machine vision-based weld seam positioning system for thermos cups is proposed. The system’s operational flow and image algorithms are designed to achieve automatic calibration of the weld seam orientation. Experimental results show a qualification rate of 97.8%, demonstrating robust performance across various cup models. The average positioning error is 45 pixels, with a maximum error of 153 pixels, corresponding to angular deviations of 0.68° and 2.30°, respectively. The average processing time of a single algorithm run is 128 ms, ensuring efficient operation in non-high-speed production scenarios. The results of this study have good application value and also provide some insights for the position calibration of other rotational objects.

Keywords

• machine vision
• algorithm positioning
• precise calibration
• automate