Research On Measurement Technology Of Large Ring Parts Base On Machine Vision

Paper topics from the Tianjin Aerospace Long March rocket Manufacturing Co.,Ltd.rocket base circularity detection needs.At present,the rocket base roundness measurement is mainly used for manual contact measurement,due to the large size of the rocket base,direct measurement of the rocket makes the measurement task heavy,low efficiency,the measurement data can not be traced,and will cause some loss of measurement tools.In view of the shortcomings in the roundness detection of rocket base,in this paper,based on the machine vision technology,the circularity measurement of large annular parts is studied.The measurement does not require direct contact with the parts to be measured.By collecting some arc images of the parts,the fitting algorithm is used to obtain the circularity,concentricity And other data indicators.The measurement method is convenient and reliable in operation,and improves the detection efficiency.It has theoretical significance and practical application value in the field of industrial measurement.Based on the machine vision related knowledge as the foundation,in view of the large circular ring parts roundness of non-contact on-line measurement,mainly from the image preprocessing,image edge detection and contour extraction and image matching are studied.Firstly,in the aspect of image preprocessing,the algorithm of gray correlation and denoising is researched and implemented.By comparing the experimental results,the best bilateral denoising algorithm is selected;In the edge detection and contour extraction step,the research and implementation of edge detection algorithm,in order to improve the measurement accuracy,based on the traditional Canny edge detection on the realization of improved Canny adaptive edge detection algorithm,which can accurately detect the edge of image,then the image edge contour extraction;In the final measurement,the size data and calculation method of ring parts are studied;In the fitting part of the image,the commonly used circle fitting algorithm is studied,and based on the traditional least squares fitting algorithm has been improved,the iterative least squares fitting algorithm and iterative fitting through the partial arc image of large circular ring parts uptake,verified the effectiveness and superiority of the algorithm.Finally,in the Microsoft Visual Studio2012 environment to develop a measurement system for large ring parts based on machine vision,and a large number of experiments for large circular ring parts,through the experimental results show that the measurement system can be measured quickly and accurately in the allowable error range,meet the requirements of the accuracy of the subject.