Research On The Rapid Measurement Method Of Fine-pitch Gear Batch Based On Machine Vision

Rapid and precise measurement is an effective means and important process to realize intelligent production of fine-pitch gear.At present,there are many equipment that can measure the geometric parameters and single deviation of gears,among which the most widely used are CMM and gear measuring center.This type of equipment is generally expensive and complex to operate,and it needs to replace the standard probe of corresponding size for the gear with different modules,plus the difficulty of alignment when installing the fine-pitch gear,resulting in low measurement efficiency.Therefore,it is very important to find a measurement method of fine-pitch gear with high efficiency,good accuracy and wide measurement range.In recent years,with the continuous breakthrough and development of machine vision key technology,machine vision has been widely used in the field of industrial measurement.In this paper,machine vision based measurement method is used to study the batch rapid measurement of fine-pitch gear.According to the demand of vision measurement of fine-pitch gear,the software and hardware components of vision measurement system are analyzed.Through the analysis and selection of light source lighting system,lens,industrial camera,optical support and other components,a hardware test platform suitable for fine-pitch gear measurement is built.The software system uses Visual Studio 2017 to call opencv4.1.0visual library,and uses 9×9 rectangular array calibration board to calibrate the fine-pitch gear visual measurement system,and obtains the scale factor between the physical size and image coordinates.The image of the fine-pitch gear is acquired by the vision measurement system,which lays the foundation for the subsequent image preprocessing of the gear and the measurement of the parameters.Two image enhancement methods,gray value transformation and Gaussian filtering,are used to highlight the gear contour information and remove the noise introduced in the gear acquisition process;OTSU image threshold processing algorithm is used to reduce the gear image data structure,improve the gear detection efficiency and obtain the binary gear image;circular mask is used to process and calculate the single gear image mask The number of teeth of gear;the extraction of image connected area can separate multiple gear images in the same field of vision to meet the needs of batch and rapid detection of fine-pitch gear;the edge data of gear is obtained by Canny edge detection operator to provide original measurement data for subsequent calculation of other parameters of gear.According to the original data collected and the characteristics of gear center point coordinates,the effective determination of gear center coordinates is realized by global least square algorithm.The accurate measurement of gear pitch deviation is realized by using the data points of gear profile near the reference circle to fit the circular arc instead of the gear involute.At the same time,other parameters of gear such as number of teeth,modulus,radius of tooth top circle,radius of tooth root circle,cumulative deviation of pitch and total deviation of tooth profile are measured.Finally,the parameters of gear are determined.The validity of the global least square algorithm and the method of using the data points of the tooth profile near the reference circle to fit the circular arc instead of the involute to calculate the pitch deviation of the gear is verified by simulation and experiment.In the actual test,the involute spur gear is taken as the measurement object,and the geometric parameters and single deviation of the same gear are measured in Gleason gear measuring instrument and the fine-pitch gear vision measurement system built in this paper.The comparison test results show that the indication error of fine-pitch gear vision measurement system is less than 7?m when measuring the tooth top circle radius,tooth root circle radius,single pitch deviation,accumulated pitch deviation,total profile deviation and other items,which can meet the measurement requirements of class 9 precision gear,and the measurement efficiency is higher than that of gear measurement center.