Research On Helical Gear Parameter Detection Method Based On Machine Vision

As a part often used in mechanical transmission,helical gear is superior to spur gear in transmission performance and load size because of its progressive tooth meshing mode,so it is widely used in industry,especially in the field of automobile manufacturing.In the traditional contact measurement process of helical gear parameters,there are disadvantages such as easy damage to the surface of helical gear,high time cost and slow measurement speed.This paper proposes a helical gear parameter detection method based on machine vision.By using dual cameras to detect the end and side parameters of helical gear at the same time,the high-precision online measurement of helical gear is finally realized.The main research contents are as follows :(1)According to the experimental needs of this paper,a reasonable image acquisition device is designed and built.In order to solve the problem that the end face camera focuses on the helical gears of different specifications,a camera lifting mechanism driven by a stepper motor is designed.In order to facilitate the subsequent extraction of a clear edge profile of the gear intermediate tooth,a rotating load platform driven by a stepper motor is designed.(2)For the detection of helical gear end face parameters,the camera calibration technology is used to calculate the pixel equivalent.In order to realize the high contrast between the contour of the end face and the background,this paper adopts a high-bright positive lighting method,and uses Gamma transform to enhance the dark part information in the end face image.Through the research and analysis of the traditional edge detection algorithm,the LOG edge detection method combining Gaussian filter and Laplacian operator is proposed to extract the contour of the end face image.The detection algorithm includes : firstly,by determining the center of the gear,the feature points on the contour of the addendum circle,the dedendum circle and the inner hole circle are extracted.Finally,the least square method is used to fit the circle to obtain the diameter of the addendum circle,the dedendum circle and the inner hole circle of the helical gear,and the concentricity error of the helical gear is detected by the fitting center of the addendum circle and the inner hole circle.The edge contour after mask processing is analyzed and counted by connected domain detection to obtain the number of gear teeth.(3)For the detection of helical gear side parameters,firstly,ROI extraction is performed on the side image to obtain the image of the middle gear tooth.The edge contour is extracted by combining the OTSU method and the Canny edge detection algorithm,and the two edge contours are linearly fitted to determine the rotation direction.The mean value of the fitted spiral angle of the two addendum circles is calculated.At the same time,considering the camera deflection problem,the edge contour of the gear boss is fitted by a straight line to obtain the deflection angle and compensate the spiral angle of the addendum circle.Finally,the spiral angle is calculated according to the conversion relationship between the spiral angle and the spiral angle of the addendum circle.(4)According to the above algorithm,a helical gear measurement system based on machine vision is designed and developed,and the visual measurement experiment of helical gear is carried out.According to the above detection technology,the helical gear parameters were measured by several helical gears of different specifications.The experimental results show that the average measurement accuracy of addendum circle,dedendum circle and inner hole circle is within 16 μm,the relative error of helix angle measurement is within0.4 %,and the concentricity error is within 0.1 mm.The measurement method proposed in this paper is feasible.