摘要
涤纶工业长丝毛丝在线检验工序一直制约着生产全流程智能制造的实现,为解决涤纶工业长丝毛丝检测问题,完善在线质量检测体系,提出一种借助机器视觉智能检测技术识别毛丝和数量统计的方法。以毛丝长度作为判断依据,基于LabVIEW图像处理技术,采用图像增强、二值化处理、数学形态学等方法获取毛丝图像,并提取长度信息。通过试验得到毛丝长度检测阈值,当图像中毛丝长度超过检测阈值就可判断毛丝的存在,同时累加毛丝数量。结果表明:该检测方案检测准确率达到90%以上,设计合理,成本低廉,对提高涤纶长丝品质和降低企业成本具有很大的实用价值。
The on-line detection process of polyester filaments hinders intelligent manufacturing in the whole production process.In order to solve the problem in inspecting polyester broken filament yarns and to improve the on-line quality detection system,a method of identifying broken filament yarns and counting the number by means of machine vision intelligent detection technology is proposed.Based on LabVIEW image processing technology,the yarn length is taken as the judgment basis.Image enhancement,binarization,digital morphology and other methods are used to obtain the filament image and to extract the length information.Through experiments,the detection threshold of filament length is obtained.When the filament length in the image exceeds the detection threshold,the existence of the filament will be recognized,and the number of such filaments can be accumulated.The experimental results show that the accuracy of the detection scheme is over 90%,the design is reasonable,the cost is low,and it has great practical value for improving polyester filament quality and reducing enterprise cost.
作者
张荣根
冯培
刘大双
张俊平
杨崇倡
ZHANG Ronggen;FENG Pei;LIU Dashuang;ZHANG Junping;YANG Chongchang(College of Mechanical Engineering, Donghua University, Shanghai 201620, China;Engineering Research Center of Advanced Textile Machinery, Donghua University, Shanghai 201620, China)
出处
《纺织学报》
EI
CAS
CSCD
北大核心
2022年第4期153-159,共7页
Journal of Textile Research
基金
中央高校基本科研业务费专项资金资助项目(2232019G-05)
国家自然科学基金青年基金项目(52103355)。
作者简介
第一作者:张荣根(1990—),男,博士生,主要研究方向为机器视觉智能检测;通信作者:冯培(1984—),女,讲师,博士,主要研究方向为机器视觉智能检测,E-mail:pfeng@dhu.edu.cn。
