Study On Manipulator Unloading Workpieces On Catenary Based On Machine Vision

With the development of industry, Industry manipulators act an import role in the area ofloading and unloading to replace manual. And the demand for manipulators is increasing all thetime. As we know, it not only needs strong labor but also causes low productivity to load orunload by hands. And it also takes too much time to recognize their characteristic in thetraditional spraying production line. According to the current situation of wheel-holespositioning and wheel handing to the line, the article will describe a new system to measure andget the position of wheel-holes. This system will finish loading and unloading without humanbeings.The work process of this system is as the following: The linear CCD camera records allthe circular images when the wheel is rotated by workbench. The system can calculate theposition of wheel-hole and control the servo motor by image processing, characteristicrecognizing and measuring to locate the position of wheel-hole automatically. Then themanipulator will take the wheel to the handing line to make sure that the hole is on the top ofwheel. The paper uses the system to locate the other wheel-hole in the similar situation. Theprocedure is as following:(1) The design of the manipulator and visual workbench. According to the requirementsof the workpiece handling, the author has designed a6DOF articulated manipulator to realizethe loading and unloading of the workpiece operation in this paper. Based on thecharacteristics of the wheel valve hole’s recognition and positioning technology, the paper hasestablished the visual measurement and positioning hardware system in laboratory. Accordingto the time of handling one workpiece, the author sets the workbench’s rotating speed and thelinear CCD camera’s parameters, thereby we can perform the image acquisition of the wheel’scircumference.(2) According to the acquisition of the wheel’s circumference, the author has determinedthe image processing scheme of the machine vision measuring and positioning system, bytheoretical analysis and experiments. The scheme includes image gray-scale transformation,morphology processing, edge detection and the recognition and positioning of the valve holefeature. And the author has determined the every step’s image processing algorithms. In Visual C++development environment, the paper adopted OpenCV which is the Intel’sopen source image processing library, to write the wheel valve hole measuring and positioningsoftware. After the image dealt with the software, we can identify the location of the valve holecenter in the image. Through calculation, the paper transmits the information to the controlsystem to control of servo motor’s rotation. So we can rotate the wheel with the platform to thedetermined location. Finally the paper has obtained the ideal recognition results, and achievedthe positioning of the valve hole.(3) The paper has performed the motion planning of the manipulator and workbench.According to the requirement of the loading and unloading workpieces, the author hasdetermined the position of the manipulator, when it’s fetching and suspending the wheel.Combining the indicators of the trajectory planning curve, the author has performed severalinterpolations’ trajectory planning for the manipulator in the joint space. It is concluded that inthe process of movement, when the acceleration function is the subsection sine function, themanipulator’s operation is relatively stable.Finally, through the experiment the paper carried out the image acquisition of the wheel’scircumference, image processing, character recognition and the measuring and positioning ofthe valve hole. The paper obtained the correct rotation angle of rotary worktable, and verifiedthe feasibility of this vision measuring and positioning system.