Motion Parameters Identification Of Yarn Based On Machine Vision

In textile production,yarn is not only the processed object,but also an indispensable part of the processing system,it is driven by mechanical parts,and its movement is large scale and large deformation,which often cause yarn deformation such as tension,torsion and bending,even accidental drafting or fracture,then affect the quality of textiles.With the popularization and application of automatic and intelligent production mode,it is particularly important to realize automatic manipulator operation such as yarn knotting and joint.Accurate measurement of yarn motion parameters and understanding the yarn motion state in production process are the basis for rational design of textile machinery and accurate regulation of process parameters.It is hard to measure yarn motion parameters with no interference and with contact sensor,and the equipment is difficult or even impossible to install.It is more suitable to measure yarn motion parameters by non-contact method.In this paper,a yarn motion parameters recognition method based on machine vision was proposed,and the main research contents are as follows:Firstly,the research status of yarn motion parameters recognition,three-dimensional measurement technology and machine vision technology at home and abroad was investigated,and the problems and difficulties faced by yarn motion parameters recognition were analyzed.Two universal yarn motion parameters test points in industry were summarized,the yarn motion parameters recognition mechanism based on limit theory was analyzed,and the yarn motion parameters recognition process and methods: image pickup and parameter recognition were determined.Then,the camera calibration method,test point location method and stereo matching algorithm in binocular stereo vision 3D measurement technology were optimized.The depth map of the image was obtained by camera calibration,epipolar correction and stereo matching,and the yarn test points were picked up by the yarn test point positioning method,so as to obtain the three-dimensional coordinates of the yarn test points.Finally,aiming at the two kinds of motion parameters test point states,the yarn motion parameters identification experimental device and program were developed.The motion parameters(trajectory,displacement,velocity and acceleration)of yarn test point were obtained by experiment and calculation.The comparison between theory and experiment showed that the experimental results were in good agreement with the theoretical results.In this paper,a method of identifying yarn spatial motion parameters through non-contact machine vision technology(image pickup and parameter recognition)was proposed,accurate yarn spatial motion parameters could lay a foundation for rational design of textile machinery and accurate regulation of relevant process parameters.