Research On Technology Of Screw Thread Parameter Measuring Algorithm And Evaluation

As one of the most commonly used connections in industrial production,The stability of equipment performance depends on the detection exactness of thread parameters,with the increased requirements for thread measurement in manufacturing technology,traditional methods of testing thread parameters are no longer adequate for the needs of modern industry in terms of accuracy and efficiency.By establishing a thread tooth distortion model and calculating the amount of thread edge distortion,and corrects the measurements of the pitch diameter and included angle.The measurement results of the parameter detection algorithm studied in this paper are compared with the results of contact measurements.The main research elements of this paper include:(1)Several methods of thread parameter inspection are discussed,and a machine vision thread parameter measurement algorithm and experimental verification scheme are given.(2)Processing of the acquired thread images,the thread image at the time of focusing is determined using the sharpness evaluation function,the image is smoothed by median filtering,the binarised image is obtained by threshold segmentation,and the thread edge is extracted by the Canny edge detection algorithm.The causes of thread profile distortion are analysed,the equation for the distorted thread profile and the equation for the true profile are derived,and the effect of the distortion on some of the parameters of the thread is analysed.(3)Using the Harris algorithm to detect angular points and the least squares method to fit the thread profile,the calculation of several parameters of the thread is obtained and the formulas for the compensation of the pitch diameter and included angle after distortion of the thread profile are derived.(4)Image calibration and comparison experiments with different sizes and types of thread plug gauges were carried out.The uncertainties affecting the results of the parameter calculations are assessed to show that the measurement algorithm is reliable.In this paper,a geometric model of thread profile distortion is established with cylindrical male thread as the measurement object,and the machine vision thread parameter algorithm and compensation formula are derived.After image calibration and comparison tests,this means that the detection algorithms and corresponding compensation formulas for each of the above parameters meet the measurement requirements.