Machine Vision Based Pattern Engraving for Linear Motor Gantry System

• Main Authors: Jun-Kai Wang, 王俊凱
• Other Authors: Chaio-shiung Chen
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/82772350567627104863

碩士 === 大葉大學 === 機械與自動化工程學系 === 102 === This thesis develops an automatic pattern engraving for linear motor gantry systems based on machine vision. A personal computer, a CCD vision system, a PCI-1240U motion card, and a linear motor gantry were integrated in the hardware structure. The programs that compose of image process, trajectories planning and motion control were written to enable automatic pattern engraving by LABVIEW software. Firstly, the CCD captured the desired pattern images. Then, image processing techniques that compose of median filter, edge enhancement, binarization, thinization, skeletonization and edge detection were designed to obtain the image coordinates of the pattern paths. These image coordinates were translated to the cutting coordinates according the dimensions of the linear motor gantry. Trajectories planning and motion control were then designed to engrave the patterns through the PCI-1240U motion card. The curvatures of the cutting paths were computed by a least square method. According to those curvatures and their change rates, an intelligent federate was developed based on a fuzzy system to achieve a high speed and precision cutting for the linear motor gantry systems. Finally, experiments were performed using a practice linear motor gantry to demonstrate the effectiveness of the proposed control methodology.