Design Of Pipe Thread Parameters Detection System Based On Machine Vision

With the continuous exploitation of oil,more and more oil pipes are put into use,which indirectly leads to an increase in the demand for pipe threads.Therefore,in order to ensure the safety of oil transportation,high efficiency and high precision requirements are put forward for the detection of pipe threads.However,the current thread detection is mainly based on contact detection,which has low detection accuracy and slow efficiency.Therefore,this article mainly studies the non-contact pipe thread parameter detection based on machine vision.On the basis of machine vision and image processing,the geometric parameters of the thread are measured,and the measured parameters are measured according to API(American Petroleum Institute)and manufacturer regulations.Perform data processing and analysis on pipe thread parameters to determine the quality of the thread.The main studies are as follows.Firstly,the detection platform of pipe thread parameters is designed and built.It includes the selection of camera and lens,the selection of lighting source and the design of servo motor motion control.In terms of software,image processing module and human-computer interaction module are introduced.Secondly,according to the processing flow,first use the Zhang Zhengyou calibration method to obtain the camera's projection matrix,distortion coefficient and perform distortion correction on the collected image.Then use digital image processing technology to extract features of pipe thread images.Specifically: use median filtering and bilateral filtering fusion filtering method to filter the thread image,use histogram operation to determine the segmentation threshold to obtain the binary image of the thread,and use the Canny operator based on bilinear interpolation to perform the edge of the image Detect and use image refinement and breakpoint defect repair processing to further extract thread edge image features.Then,using the standard gauge block of known size,the digital image processing technology is used to collect and calculate the gauge block,and finally the ratio coefficient between the actual size of the gauge block and the pixel size of the collected image is obtained.According to the distribution characteristics of pipe thread,the progressive scanning feature point algorithm,Floyd(Insertion point method)based pitch diameter extraction verification algorithm and RSG(Region Seeds Growing)algorithm are used to obtain the corresponding feature point information of pipe thread tooth height,pitch,taper,crown height and profile angle before fitting with the existing least square method,and the corresponding pipe thread parameter measurement algorithm is proposed according to the characteristics.Finally,the control analysis of the entire detection system is carried out,and human machine interface of the system is introduced in detail.Use the 60 sets of pipe thread data to perform repeatability verification,calculate the maximum value,average value,and limit error of each parameter of pipe thread respectively,and compare with the standard parameters provided by the manufacturer to verify the feasibility of the detection algorithm proposed in this paper.After that,the error in the system is analyzed,and the random error in the system is solved by calculating the absolute error and relative error of the thread parameters,and the system error is solved by algorithms such as fusion filtering.