中国烟草学报2025，Vol.31Issue(4)：22-33,12.DOI:10.16472/j.chinatobacco.2024.T0314
基于3D视觉的GDX6包装机商标纸智能装载系统
Intelligent loading system for trademark paper of GDX6 packaging machine based on 3D vision
摘要
Abstract
In response to the issues of low efficiency and high cost of manual loading of trademark paper in the GDX6 packaging machine,this paper designs an intelligent loading system for the trademark paper based on 3D vision.The system uses a 3D vision camera to capture trademark stack images.The YOLOv5s-seg instance segmentation model is used to detect the trademark stacks.The detected flat coordinates of the trademark stack are matched with the scene point cloud information to obtain the segmented point cloud data,ultimately determining the pose of the trademark stacks.According to the pose results,a collaborative robot delivers the trademark stacks to a non-destructive cutting and separation device to perform non-destructive cutting and separate the wrapper paper.Subsequently,the trademark stacks are transported to the packaging machine conveyor belt through a placing material storage.The results show that the system can quickly perform instance segmentation and pose estimation of the trademark stacks,achieving the requirements of normal production processes and replacing manual labor.This improves the automation level and production efficiency of the packaging machine unit.关键词
GDX6包装机/实例分割/3D视觉/协作机器人Key words
GDX6 packaging machine/instance segmentation/3D vision/collaborative robot引用本文复制引用
张志彬,蒋景强,陈瑾龙,张小杭,王己锋..基于3D视觉的GDX6包装机商标纸智能装载系统[J].中国烟草学报,2025,31(4):22-33,12.基金项目
福建中烟工业有限责任公司技术改进类重点科技项目"高速包装机商标纸智能装载系统自主研发与应用"(FJZYKJJH2023ZD036) （FJZYKJJH2023ZD036）
