摘要
针对瓦楞纸箱表面缺陷对纸箱质量的影响问题,建立了纸箱表面缺陷数据集,设计了基于YOLO v3算法对纸箱表面缺陷进行检测的深度神经网络模型,并使用PyQt工具设计了能够识别纸箱表面缺陷的软件,它与人工检测相比具有更高的速度、精度和稳定性,能够实现瓦楞纸箱表面缺陷特征实时检测。
This paper describes a solution for detecting surface defect in cardboard cartons based on a deep neural network design.The approach is to use a neural network to recognize the characteristics of surface defects in real-time and provide more accurate,faster,and stable detection compared to manual detection methods.
作者
吴飞
蒋罗彬
贾伟萍
刘文婷
沈大伟
杜刚
WU Fei;JIANG Luobin;JIA Weiping;LIU Wenting;SHEN Dawei;DU Gang(School of Mechanical and Electronic Engineering,Wuhan University of Technology,Wuhan 430070,China;China Tobacco Hubei Industrial.LLC,Wuhan 430040,China)
关键词
瓦楞纸箱
缺陷检测
机器视觉
深度学习
神经网络
corrugated box
defect detection
machine vision
deep learning
neural network
作者简介
吴飞(1973-),男,湖北武汉人,武汉理工大学机电工程学院教授.
