| With the development of image processing and machine vision technology, industrial field have introduced vision technology, such as intelligent service robots, intelligent industrial robots, EOD robots and so on. This is greatly improved the production efficiency and saved time and cost. It has very important significance to the development of industrial technology. However, there are still many problems when the visual technology applied in the actual industrial production, such as poor production conditions and poor adaptabilities, etc. These all need to constantly improve the accuracy and efficiency of visual identity to be effectively solved.This paper studies the industrial vision robot and sets up a visual crawl robots platform. The system adopts RH6 industrial robot from XinSong Company as body, tie-in industrial camera and image processing procedures, to identify grab workpiece tray artifacts within the operations.This article firstly design and analysis system overall, based on Linux system structures, set up the communication link between the related equipment. The PC machine is the server side, and the robot controller is the client. Secondly, research the industrial camera calibration accuracy. Based on OpenCV library, write the calibration procedures. Comparing with Matlab toolbox calibration results, and through the actual experiment, proved the calibration procedure is effective and feasible in this paper. Last, tests of different shapes of the workpiece are classified, puts forward a method of rapid detection of the workpiece center of mass. The target image is decomposed into pieces, using hough transform to detect type of circular workpiece, the angular corner features combined with torque testing object, and through the C language programming.Based on the above study, build the actual robot vision system. Through the actual condition test, prove that the design method of this paper is feasible. According to the method proposed in this paper, improve the identification precision of the workpiece and the computing speed of the program. |
