The Research Of Image Processing Technology In The Manipulator Drawing System

With the rapid development of machine vision technology,industrial robots with visual guidance systems have been rapidly promoted.At present,industrial equipment such as laser cutting robots and welding robots,which are generally used,are often used to manually input workpiece information before the workpiece is cut or welded.A large workpiece needs several hours to complete.The development of visual technology has brought forward a new and quick solution for it.It uses visual equipment to shoot and record the workpieces,analyze the workpiece information,and convert it into workpiece data that needs to be manually entered.This paper uses machine vision to assist robot motion control in drawing human faces and signature handwriting to verify this theory:camera capture captures portraits and handwriting,uses image processing,vectorizes bitmaps,maps pixel coordinates to world coordinates,and finally plots the results by robots.On paper.In this paper,when implementing this solution,the PCI motion control card is not used for the robot motion control system.The use of soft motion control cards(software programmed motion control programs)and industrial bus control robots based on the EtherCAT protocol.More scalable,lower cost,and more real-time.The main work of this paper is as follows:(1)Use an industrial camera to take pictures of faces,and use the Local Adaptive Contrast Enhancement algorithm to enhance the key image information such as face contouring.At the same time,the integral image is used to binarize the image,which reduces the effect on the image processing result caused by the uneven illumination due to the uneven illumination in the daily environment.After obtaining the face contour curve in the face photo,in order to ensure the stability of the robot during the drawing process,the profile curve is anti-aliased and burrs,and finally the pixel coordinates are mapped to the actual physical coordinates to draw the human face.(2)Compared with the conventional image refinement algorithm,the acquisition of handwriting signatures is based on the analysis of the axis of the image and the topological structure of the signature.(3)The image processing program uses Visual studio 2015 combined with Qt,with OpenCV(Open Sourc-e Computer Vision Library),written in c++.Using the CoDeSys(Controlled Devel-opment System)compiled ST(Structured text)language program,the robot motion control is performed in the CoDeSys RTE real-time kernel program.and use inter-process communication to achieve two Commands and data exchange between different programs.