Research On Key Technology Of Computer Sewing Machine Based On Machine Vision

In clothing and shoes manufacture, there is an urgent need of flexibility of equipment due to various sizes, unstable cutting and pliant materials etc. On basis of the traditional machine, this paper improves and designs a computer sewing machine which is based on machine vision, which utilizes digital image processing techniques to extract the edge contour of cutting pieces on the basis of images of the cloth and shoes samples captured from camera, and then sews according to the processing information of patterns. The device is a more intelligent and automatic computer sewing machine. Main work of this paper is as follows:By studying of the whole control process, PC with a touch screen and motion controller as the control scheme is determined. And the hardware platform of the control system is analyzed in the paper. Meanwhile, as the machine vision system is of extreme significance, this paper analyzes the hardware of image acquisition module selection, mainly including the selection of industrial camera, lens and light source. Besides that, the camera calibration method that using the Open CV library functions calculates the camera parameters and the distortion coefficients is studied with the purpose of eliminating the negative effects of camera distortion imposed on images.Researching on the key algorithm of the edge detection, firstly, analyses image filtering processing. After two kinds of edge contour extraction algorithm based on the gradient operator and threshold segmentation are analyzed and compared, Otsu method is chosen to segment images. Then study on morphological processing of binary images, fulfilling extraction of the edge contour of cutting pieces.In the software development and designing, the paper discusses several different kinds of data files, elaborating the process of extracted edge contour stored as PLT files, of storing the pattern graphic as SDT files. Additionally, the process of image acquisition and image processing program has been expounded. Researching the curve fitting with B-spline specifies to fit irregular graphics. Analyzing the discrete method of straight line and B-spline to implement contour curve discretization generates data files involving stitch length and stitch coordinate processing information. Finally, in the paper the display process of pattern graphics has been analyzed.