Font Size: a A A

Research On The Calibration Of Mounting Machine Vision System

Posted on:2015-04-19Degree:MasterType:Thesis
Country:ChinaCandidate:S H ZhangFull Text:PDF
GTID:2298330422990970Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Precision placement machine is the core electronics manufacturing industry, and isthe flag of modern scientific and technological strength. Precision placement machinestructure is complex, involving multiple disciplines, electronic, mechanical, softwaredesign, optics, etc. The visual system is the core part of SMT. This article will describethe placement machine vision system calibration theories and programs to improve themeasurement accuracy. By compensating for the visual system, the control precisionwill meet the requirements of new generation of IC chip. This paper focuses on theplacement machine vision system calibration technology, the main contents are:(1) Parameter of the placement machine vision system calibration is required,including the visual system’s internal parameters and mechanical parameters. Internalparameters caused the deviation between the position of the pixels in the image and theactual point. Mechanical parameters are the installation error of the visual system, andthe installation error gives calculated results a presence offset. In order to improve theproduction of precision placement machine we are required to obtain two parameters.(2) Internal parameters of placement machine vision system can be obtained fromthe camera calibration technology. This paper analyzes the linear and nonlinear modelsto get the parameter which is needed by camera calibration. An optimization method isproposed based on Tsai’s method and Zhang’s method. We can to obtain the exact resultof camera parameters by solving a nonlinear optimization problem.(3) Circular feature point detection and location technology is an image processingtechnology. It can provide position reference for us to extract mechanical parameters ofvisual system. The general flows of feature point detection technique are edge detection,Hough transform, sub-pixel edge processing, least squares fitting edge. Ultimately, wecan get the exact coordinates of feature points. This article compares tracking andgradient detection two kinds of edge detection techniques for edge detection. UsingHough transform technology RHT improved we can improve computing speed.(4) Mechanical parameters of the visual system, including the deflection angle andoffset of the camera. By designing the positional relationship between the camera andthe feature points, we can get the deflection angle and offset of the visual system.
Keywords/Search Tags:Placement machine, Camera calibration, Nonlinear optimization, Featurepoint detection, Sub pixel
PDF Full Text Request
Related items