应用YoLov5的网布质量检测系统设计  

Design of Mesh Quality Detection System Using YoLov5

作　　者：王嘉铭 高张慧 江晓蕾 张子豪 马梓博 WANG Jiaming;GAO Zhanghui;JIANG Xiaolei;ZHANG Zihao;MA Zibo(KeWen College,Jiangsu Normal University,Xuzhou,China,221100)

机构地区：[1]江苏师范大学科文学院,江苏徐州221100

出　　处：《福建电脑》2024年第12期14-19,共6页Journal of Fujian Computer

基　　金：江苏省高等学校大学生创新创业训练项目(No.202413988016Y)资助。

摘　　要：为改善人工检测在纺织质量检测域中效率不足的问题,本文设计了一款基于深度学习技术的网布质量检测系统。系统应用CNN卷积神经网络模型和深度学习YoLov5算法构建,采用工业相机捕捉网布图像,对提取的特征进行检测,并可视化呈现数据分析结果。测试的结果表明,系统检测准确率达85.7%,检测速率达74.1Hz。该系统可应用于中小型企业进行网布材质的品质检测。To improve the efficiency of manual inspection in the field of textile quality inspection,this paper designs a mesh quality inspection system based on deep learning technology.The system is constructed using CNN convolutional neural network model and deep learning YoLov5 algorithm.Industrial cameras are used to capture mesh images,detect the extracted features,and visualize the data analysis results.The test results show that the system has a detection accuracy of 85.7%and a detection rate of 74.1Hz.This system can be applied to small and medium-sized enterprises for quality inspection of mesh materials.

关 键 词：网布材质 质量检测 深度学习 YoLov5算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]