The Design Of Automatic Positioning Punching Machine Based On Machine Vision

The precision of punching can directly affect the precision of product. The AutomaticPositioning Punching Machine can improve the precision of positioning hole by virtue ofAutomatic Positioning and Vision System. So more and more companies introduceAutomatic Positioning Punching Machine as a special punching equipment. The machinerepalces the traditional manual punching and improve the quality of product. They alsoimprove the competitiveness of enterprises. For this reason,we need to design ahigh-precision Automatic Positioning Punching Machine to meet productionrequirements.Machine vision is a comprehensive technology, including vision sensor technology,lighting technology, optical imaging, digital image processing technology, analog anddigital video technology, computer software and hardware technology and automaticcontrol technology. In the production line, machine vision technology can be used forobject identification, position detection, completeness checking, shape and dimensionalinspection, surface inspection. The image analysis and processing algorithm in machinevision is the core technology, including edge detection, template matching, characterrecognition. And template matching algorithm can solve the object location, calibration,measurement and detection. It plays an important role in machine vision.This paper expounds the developing situation and application of the machine vision,the composition of machine vision systems. As the algorithm theory of automatic drillingmachine system design, this paper summarizes some template matching algorithm, such asGray-Value-Based Template Matching, Matching Using Image Pyramids,Subpixel-Accurate Gray-Value-Based Matching, Template Matching with Rotations andScalings, other Robust Template Matching.Based on deeply analysis of the template matching algorithm, this machine visionalgorithm is applied to the Automatic Positioning Punching Machine. Through template matching algorithm, We process the CCD camera view range images and completesample comparison and confirmation, in order to get the correct object and achieve thehigh-precision positioning. We use the general machine vision software-Matrox ImagingLibrary, developed by Matrox company in Canada. The software technology is matureand have rich function. Generally, the geometric matching algoriyhms have an advantageif the template and image contain only few, salient geometric primitives. So, we use thegeometric matching algorithms in this automatic system. The hardware of systemconsists of image acquisition and motion control,and motion control is divided intocontrol, transmission, pneumatic three parts. Through the computer, it capture and analyzethe sampling image from the camera to control the motion parts. The final result is thesystem position error less than0.008mm, drilling speed less than0.4s per hole, meet thesystem requirement of positioning accuracy and drilling speed.