机器视觉在卷烟生产中的应用与展望  

Application and Prospect of Machine Vision in Cigarette Production

作　　者：李郸 梁洲源 马慧宇 李海燕 王春琼 张轲 廖泽容 张榆锋[2] LI Dan;LIANG Zhou-yuan;MA Hui-yu;LI Hai-yan;WANG Chun-qiong;ZHANG Ke;LIAO Ze-rong;ZHANG Yu-feng(Yunnan Tobacco Quality Supervision and Inspection Station,Kunming 650104,China;School of Information Science&Engineering,Yunnan University,Kunming 650500,China;School of Rehabilitation,Kunming Medical University,Kunming 650500,China)

机构地区：[1]云南省烟草质量监督检测站,云南昆明650104 [2]云南大学信息学院,云南昆明650500 [3]昆明医科大学康复学院,云南昆明650500

出　　处：《计算机技术与发展》2024年第1期213-220,共8页Computer Technology and Development

基　　金：中国烟草总公司云南省公司科技计划重大项目(2022530000241036);国家自然科学基金(6226010174);云南省科技厅基础研究计划(202201AY070001-035)。

摘　　要：烟草是国家和地方财税的重要来源,而卷烟是最主要的烟草制品。卷烟生产主要包括烟叶加工、制丝工艺、烟支卷接和卷烟包装四个环节,针对卷烟生产工艺中传统人工或半自动方法耗时费力且易造成较大误差的问题,大量基于机器视觉的方法与应用被提出,进一步提高卷烟生产自动化和标准化程度。为深入了解基于机器视觉的卷烟生产研究,整理和分析了近年来国内外的相关文献。首先,阐述了机器视觉基本原理和卷烟生产的主要工艺流程;然后,从图像处理技术、传统机器学习和深度学习三类,对近年来机器视觉在卷烟生产中的研究与应用进行梳理,并分析了现有方法面临的挑战;最后,对机器视觉技术在卷烟生产领域的应用进行展望。该文系统梳理了机器视觉在卷烟生产中的应用,对卷烟的人工智能化生产具有重要参考价值。Tobacco is a significant contributor to the fiscal revenue of both national and local governments,and cigarettes are the predominant tobacco product.Cigarette production involves four main stages:tobacco leaf processing,cut tobacco processing,cigarette rolling,and cigarette packaging.Traditional manual or semi-automatic methods in the cigarette production process are inefficient,laborious,and error-prone.Therefore,a plethora of machine vision-based methods and applications have been developed to enhance the automation and standardization of cigarette production.In order to further understand the research of cigarette production based on machine vision,the relevant literatures at home and abroad in recent years are collated and analyzed.Firstly,the fundamental principle of machine vision and the main process flow of cigarette production are introduced.Secondly,the research and application of machine vision in cigarette production are categorized into three groups:image processing technology,traditional machine learning,and deep learning,and their advantages and challenges are discussed.Lastly,the prospects of machine vision technology in the field of cigarette production are presented.We offer a systematic overview of the application of machine vision in cigarette production,which has important implications for the intelligent production of cigarettes.

关 键 词：机器视觉 卷烟生产 图像处理 机器学习 深度学习 人工智能 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]