Research On Monitoring Platform Of Intelligent Sorting And Feeding System

In recent years,with the continuous expansion of production scale,the continuous increase of labor cost and the adjustment of national development strategy,the textile industry needs to continuously improve its level of automation and intelligence.The emergence of intelligent sorting and feeding system of industrial robots can effectively replace the manual process of yarn defect detection and sorting and feeding.The importance of this system is not to mention.And metaphor.Based on this system,this paper designs a monitoring platform for intelligent sorting and feeding system.The main tasks are as follows:(1)Design the overall scheme of the monitoring platform.Select Linux operating system and ARM as the software and hardware platform of video capture terminal,Windows as the development platform of video receiver;then analyze S5PV210 processor,camera,Ethernet interface and mechanical arm,and finally design the software scheme.(2)Three kinds of yarn detection algorithms,i.e.inter-frame difference method,standard ViBe algorithm and improved ViBe algorithm,are experimentally compared and their advantages and disadvantages are analyzed.The experimental results show that the improved ViBe algorithm not only has the advantages of good real-time performance,fast update speed and strong stability of background model,but also can detect the yarn area completely,and is more suitable for the research of this subject.Finally,HSV color segmentation is incorporated into the improved ViBe algorithm,which can realize the color recognition of yarn captured by the manipulator very well.(3)Design and development of monitoring platform applications.The application of the platform mainly includes three parts: embedded Linux system porting,server programming and client programming.This topic uses V4L2 video driver interface,H.264 hard coding and UDP communication protocol to realize video data acquisition,encoding,transmission and reception,etc.,using robot arm API,OpenCV,database to complete the control of the robot arm,video image display,algorithm Processing and data logging.Finally,the monitoring platform is tested as a whole.The test results show that the monitoring platform can not only quickly identify the color of the bobbin captured by the robot arm,but also meet the requirements of the design of the subject in terms of video data acquisition,robot arm debugging,and database recording.