朱鹏 倪俊芳

摘要：缝编机工作时，氨纶丝纤细透明一旦断经，未能按时报警停车，就会织出废品。而传统人工目检和机电断经检测系统费时费力，且易产生误判。基于Windows和C++平台开发了机器视觉检测系统，采用摄像机间隔拍摄经编图片，通过视觉算法对采集图片处理分析，判断断经与否，并通过USB4761通信口通断控制缝编机的启停。实际应用表明该系统对于断经判定的稳定性得到提高，报警率达到100%。

关键词：缝编；经丝；视觉检测；C++

中图分类号：TS103.7文献标志码：A文章编号：1009-265X（2018）03-0093-04Research on Broken Warp Detecting System of

Stitching Machine Based on Visual Detection

ZHU Peng， NI Junfang

（School of Mechanical and Electrical Engineering， Soochow University， Suzhou 215021， China）Abstract：spandex warps， which are slender and transparent， will be woven out of scrap during working of the stitching machine if warps once disconnected and the alarm failed to work timely. Because the traditional manual inspection and electromechanical detecting system are timeconsuming， laborious and easy to result in false alarm， the detecting system of machine vision is developed based on windows and C++ platform. Images are shot through the camera at intervals of 2 seconds， processed and analyzed by vision algorithms to judge if the warp is broken， then that the stitching machine works or not will depend upon the onoff signals of USB4761 communication port. The industrial applications show that the judge stability for the warp broken or not is improved， the correct alarm rate reached 100%.

Key words：stitch；warp；visual detection； C++

通信作者：倪俊芳，Email：jfni9999@sina.com缝编机是一种基于经编原理的编织机械，氨纶丝由经轴引出，经过穿丝板、导纱针，垫入槽针，针芯与槽针不间断前后反复运动，对非织造布穿插，导纱针在花型凸轮的带动下横移，将缝编纱喂入，形成缝编织物[1]。缝编机生产弹力布过程中，每台缝编机有近千根氨纶丝，其纤度极小仅为77.8 dtex（70D），透明且有弹性，一旦断经，织出废品，必须停车裁剪，重新拼接布料，接丝开机，费时费力。现有的断经检测系统虽然可以在断经后发出报警信号并停车，但也存在很多的缺点，比如安装近千个停经片及穿经工作量极大，不能快速确定断经位置，此外，电控停车装置的功能还不够完善等。目前，国外采用机器视觉技术检测经丝断裂技术上已成熟，国内采用嵌入式智能相机检测经丝断裂获得成功[2]，但是一套断经检测系统成本很高。因此根据企业实际需求状况，本文将采用普通工业相机采集图片，编写C++程序通过视觉算法检测断经与否。……