初生纤维根数在线识别系统研究  

Online identification of the number of the as-spun fibers

作　　者：李飞 冯培[1,2] 杨崇倡[1,2] 张俊平 马勋勋 刘大双 LI Fei;FENG Pei;YANG Chongchang;ZHANG Junping;MA Xunxun;LIU Dashuang(College of Mechanical Engineering,Donghua Universily,Shanghai 201620,China;Engineering Research Center of Advanced Textile Machinery,Ministry of Education,Shanghai 201620,China)

机构地区：[1]东华大学机械工程学院,上海201620 [2]纺织装备教育部工程研究中心,上海201620

出　　处：《上海纺织科技》2023年第2期62-64,共3页Shanghai Textile Science & Technology

基　　金：国家自然科学青年基金项目(52103355);国家重点研发计划资助(2017YFB1304000)。

摘　　要：为了降低化纤长丝生产中对纤维飘丝、断丝等的人工巡检强度和频次,提出了一种基于机器视觉的初生纤维根数在线检测方法。为准确获取初生纤维丝的图像,搭建了点光源照明的图像采集平台;在化纤生产过程中在线进行图像采集,利用滤波、二值化处理计算得到初生纤维根数;开发了适于Windows系统和Android系统的初生纤维根数在线识别系统的人机交互界面。在生产线上的应用结果表明:该系统提高了在线检测的效率和准确率,降低了工人的劳动强度。In order to reduce the intensity and frequency of manual inspection of the as-spun fibers produced by chemical fiber filaments,an online detection method for the number of the as-spun fibers is proposed according to the image processing techniques.In order to accurately acquire the images of the as-spun fibers,an image acquisition platform illuminated by a point light source is established.The image of the as-spun fibers on the cross-section in the spinning production line is acquired via the platform,and the number of the as-spun fibers is obtained through filtering and binarization.The man-machine interface of the online identification system of the as-spun fibers suitable for Windows system and Android system is developed.The application results on the production line show that the system improves the efficiency and accuracy of detection and reduces the labor intensity of workers.

关 键 词：初生纤维 机器视觉 在线识别 图像处理 

分 类 号：TS103.6[轻工技术与工程—纺织工程]