Research On Machine Vision System For Punching Mark Of Textile Upper

The punching of the textile uppers’ marks is an important process in the stream-lined line of shoes after the upper printing,which determines the accuracy and precision of the positioning,and affects all the subsequent shoe making processes.However,most domestic shoe-making enterprises adopt the traditional manual punching process,which compared with the developed countries who adopt machine vision technology to make punching automatic,its inefficiency is self-evident.In recent years,with the rapid development of computer technology and information processing technology,machine vision technology gets the comprehensive popularization in China,especially the gradual deepening of industrial manufacturing field which is the only way to realize automatic and intelligent production.The application of machine vision technology to the mark punching of the textile uppers can solve the disadvantages of the traditional manual process and also improve the production efficiency and the punching precision.The purpose of this paper is to design a machine vision system to realize the automatic punching of the textile upper marks.The system mainly includes camera calibration,image pretreatment and the extraction of punching marks.Firstly,calibrate the camera used and the reconstruct the image to offset the distortion.The calibration results directly affect the positioning accuracy.Then,preprocess the image of shoe pattern,such as graying,filtering,binary etc..Finally,extract the punching marks and save the coordinate information.In this paper,two schemes are designed to extract the punching marks,which can be divided into "hit and miss" scheme and template matching scheme.Both of the two schemes can successfully extract the location information of the punching marks.Both of the two schemes has advantages and disadvantages,so the choosing of schemes should be according to the actual situation of the shoe pattern.In this paper,the software of HALCON is used to test and simulate the scheme.The function of HALCON is so comprehensive that the vast majority of image processing can be achieved by operator,which make the whole process simple and reliable.The weighted median filtering technology is used to process image,which protects the edge information of the image to a certain extent and helpful to the template matching operation that is characterized by the outline of the punching marks.For feature based template matching,HALCON also provides a powerful matching operator,simplifying the workload of the researchers.Both of the two schemes designed in this paper achieve better test and simulation results in the HALCON.The research on machine vision system for punching mark of textile upper originate from the production line of shoe-making enterprises.The research is the actual demand of the shoe-making enterprises which had a certain practical application value.