Research On Deviation Detection Technology Of Cloth Printing Based On Video Processing

The printing deviation detection in the automatic printing process is an important technical guarantee for cloth printing performance,and it is an important reference for the adjustment of the printing position of the servo control system.The detection method of manually distinguishing the printing deviation cannot meet the speed and quality requirements of control and adjustment.The technique of detecting the printing deviation of cloth based on video processing can solve this problem very well.In the actual cloth printing process,a cross mark is usually set in front of the printing area to locate the beginning of the printing area.The deviation of each process is given by the subsequent multiple circular marks and the number offset in them.In this context,a large-field-of-view camera is used to locate the crosshairs,and a small-field-of-view camera is used to capture the image of the printed area.The circular deviation and the digital center are used to detect the deviation of the printing.The main tasks are as follows:First of all,in this paper,the method of locating the center of the crosshairs under dual fields of view is used to effectively collect the printed images of the production line,and preprocessing is performed by means of cropping,filtering,edge detection and tilt correction.Secondly,a method of detecting deviation based on detecting the center of a printed circular pattern is proposed.Highlight the circular features of the image through edge detection,and the deviation value of the printed pattern is obtained by using Hough circle detection and least squares circle fitting.Then,another printing deviation detection method based on the digital center position detection in the printing circle is proposed.The numbers of the collected printing images are marked out by a convolutional neural network,and the smallest bounding rectangle of the number is obtained in the marked box and the center position is calculated.In addition,the process numbers in the circle are identified by a simple projection distinguishing vector.Finally,by comparing the measured values with the actual values of the two methods,it is concluded that the cloth printing deviation detection result based on the digital center in the circle has a smaller detection error and better detection stability.And an embedded cloth printing deviation detection system based on the digital center in the printing circle is designed.