Research On Onsite Calibration And Precise Locating Technology Of Image Measurement System

With the rapid development of machine vision and image processing technology,image measurement technology has become an issue of current research on new comprehensive measurement system,due to its high accuracy,excellent real-time performance,wide application range and non-contact measurement method.This article has researched its application in measuring static parameters of the inner space,and the research points mainly include the onsite calibration and precise locating technology.Firstly,straightness,radial size and volume are taken as examples to introduce image measurement principle of static parameters of the inner space,then an experimental platform is set up,which is based on a computer,measuring target,optical imaging module and motion controlling module.The working mode and the function of software modules are introduced in general,and the details and difficulties of measurement process are analyzed according to the research content of this article.Secondly,taking the target image of the inner diameter measurement experiment as an example,pre-processing work is performed,according to the degradation phenomenon of image's quality.An image noise removal method based on the Total Variation model is introduced,the piecewise nonlinear gray-scale transformation is applied to the noisesuppressed image,and the edges of reticles are sharpened.These processes can effectively enhance the reticle object of the target image,which facilitates precise locating.Thirdly,aiming at the distortion of target image,imaging model and distortion model are constructed,relevant camera parameters and distortion coefficients can be obtained by using Zhang Zhengyou calibration method,and the rectification of distorted image is achieved with the help of pixels' gray-scale interpolation.Based on this,a method based on Extended Kalman Filtering is introduced,its accuracy and real-time performance can meet the requirements of the measurement system in onsite calibration.Finally,for reticle locating,related theories of image segmentation using threshold and line fitting as well as localization are studied,and a segmentation method effectively separating reticle and background of target image is introduced,it is based on the top-hat transformation which is a morphological approach.After thinning process,line detection using Hough Transform is applied to fit and locate the skeleton of reticle,experiments show that the error does not exceed 5?m,and thus accurate locating of the reticle is achieved.