Research And Implementation Of Visual Positioning Technology For Automatic Plug-in Machine

With the implementation of China made 2025 strategic plan,there are more and more products and devices for intelligent manufacturing in China,such as the automatic plug-in machine which the production facility could automatic assembling electronic components.In the automatic plug-in process,the mark center coordinates of printed circuit board(PCB)are used as reference points,the plug-in mechanism picks up different components and accurately inserts them into corresponding workstations.In this paper,the central coordinate information of the PCB positioning hole is accurately obtained by using the machine vision positioning technology.The experimental results show that the average positioning accuracy of the algorithm described in this paper is 0.1pixels on the image and the actual positioning accuracy can also reach 0.01 mm,and the running time of the system is about 320 ms,which satisfies the design requirements of the automatic plug-in machine.The contents of the research are as follows:1.We study how to use image matching technology to locate the target area and extract it to processing in later.The template matching algorithm based on gray value and local feature matching algorithm are emphasized.The performance and characteristics of these two algorithms are tested and analyzed.2.On the basis of rough positioning,the image preprocess technology is used to accurately obtain the contour edge of the positioning hole,so as to prepare for the precise positioning in the later.The preprocessing mainly includes five main contents,such as image denoising,threshold segmentation,morphological opening operation,edge detection and contour tracking.Firstly we need to smooth the target area for noise suppression;then use the adaptive threshold segmentation algorithm to separate a circular positioning hole and the surrounding background;Secondly we use morphological opening operation to remove unnecessary background and convex region in the image,so that the positioning hole edges is more clearly;Thirdly,we use Canny algorithm to detect the image to obtain single pixel edge contour;Fourthly,using contour tracing technique to sort these pixel points,and then calculate the area and perimeter of each contour;Lastly setting the roundness threshold to extract the contour of the positioning hole.3.When improving the positioning accuracy of the system,it is not only to improve the resolution of the hardware,but also the positioning accuracy of the software needs to be considered.To improve the location accuracy of contour edges by sub-pixel edge detection algorithm,so that the accuracy of the whole system can be improved.The principle of sub-pixel edge detection is emphasized,and the advantages and disadvantages of interpolation method,fitting method and moment method are compared and analyzed according to the edge morphology types of location holes,and then a detection algorithm suitable for edges of the PCB mark is selected.4.The circular positioning technique is used to accurately locate the sub pixel edge points obtained.The principle and algorithm steps of Hough circle transformation and robust least squares fitting circle method are mainly introduced.The experimental results of two algorithms are analyzed,and compare their location accuracy,robustness and efficiency.the best algorithm is picked up to embed into the system.5.Developing the visual positioning system of the automatic plug-in machine and testing the modules function of the system.The Visual Studio 2015 MFC frame is used to design the user interface of the visual positioning system and to program the communication code of the visual positioning system to the plug-in control system.