Development Of Visual Inspection System For Microstructure Intensive Array Punching Machine

LCD booming in recent years,mainly used in mobile phones,digital cameras,tablets and other digital products.As the display shows quality requirements continue to improve,light guide microstructure precision manufacturing technology requirements are getting higher and higher.In order to solve the micro-structure of the light guide plate ultraprecision manufacturing research problems,to achieve processing diameter range of 35-55 micron dense array of small circle,the processing accuracy of 3 microns within the target.In this paper,the demand analysis of microstructure-intensive array processing machine tool is analyzed,and the research on the development of machine vision detection system and the voltage-displacement curve of piezoelectric ceramics are put forward.The following is the main research content of this subject :The first is the image sub-pixel edge extraction.Using Zernike 7 ?7 convolution kernel rough detection of sub-pixel edge points,and through the experimental and theoretical analysis 7 ?7 convolution kernel than 5?5 convolution kernel more about edge direction,more detailed investigat ion.Then,on the result image of Zernike test,this topic design of eight templates to find small circle the edges of the real.The four templates are precisely positioned for the intersection of the small circle and the coordinate axis,the other four templates set out the “carpet” search for the remaining edge points of the small circle from the intersection point,thus realizing the sub-pixel edge detection of the small circle.The second is the camera calibration.This subject uses the quantum behavior particle swarm optimization algorithm(QPSO)to calibrate.In view of the characteristics of small camera depth and narrow field of view,this paper presents a initial value subjective estimation based on QPSO.This method is a scientific and effective estimate of the principle model of each parameter of the camera,its advantage is no calibration,without complicated algorithm,and has good convergence.Finally,the piezoelectric ceramic voltage-displacement characteristics research.Design the two groups of experiments,the first group: the study of fixed voltage change rate under the voltage-displacement relation curve;the second group: the study of different fixed voltage change rate under the voltage-displacement repeat output characteristics.The experimental results show that the fixed voltage change rate within the 100 times per second,rising voltage curve of different peak voltage and fixed the variation of the voltage,the voltage-displacement relation curve has a good repeat output characteristic.In summary,this topic on the basis of previous studies,put forward the precise subpixel edge detection algorithm and the initial value of subjective estimation method,greatly reduces the cost of scientific research,achieved in testing techno logy.The detection system detection average error is 0.67 ?m,has the characteristics of the ease of use and reusable.And then combined with the visual inspection system with piezoelectric ceramic voltage-displacement relation curve pattern,which solves the scientific research problems of the tool can not be adjusted precisely before machining.It greatly improves the processing efficiency and realizes the machining error of certain small circle diameter less than 3 ?m goals.These studies have brought very important use value and practical significance to the development of microstructure-intensive array processing machine tool.