| Spiral bevel gears play an important role in mechanical industry.They are widely used.In the process of machining the spiral bevel gears,it is necessary to measure the assembly spiral bevel cutter in order to ensure the quality of the spiral bevel gears and the life-span of milling cutter.According to the measured result of radial runout and high runout adjust the assembly spiral bevel cutter.The blade located on the cutter is measured by contact in recent years.Traditional measuring method used the dual-probe to measure the radial runout and high runout.This method is low efficiency and precision of measurement.Because of the contact force influence,the errors of measurement are added.CCD camera,camera calibration,image processing and feature extraction are used to measure the assembly spiral bevel cutter in non-contact method.This method eliminates the impact of contracted force and improve the efficiency,also it could avoid the abrasion of cutter.The main contents are as following:(1)The hardware of image acquisition is designed.It include CCD camera,LED light source,method of light irradiation.(2)According to the calibration theory and characteristics of the assembly spiral bevel cutter,the traditional camera calibration method is used in this paper and carried on the preliminary experiment.(3)Through the study of a variety of image processing methods and combined with the actual situation,appreciate methods of image preprocessing are used to this paper.This paper improves the traditional Canny operator,so the precision of edge detection is enhanced.After edge detection,extract effective edge.(4)Analyzed a variety of the sub-pixel edge detection algorithm,this paper use the improved sub-pixel edge detection algorithm based on Gauss fitting.Studied parameters’ calculation methods of the assembly spiral bevel cutter.It realize the measured function of the assembly spiral bevel cutter’s parameters.This paper completes the experimental platform.It finish the camera calibration and improve the traditional Canny operator and improved sub-pixel edge detection algorithm based on Gauss fitting.Through the experiments,the method is feasible and realize the non-contact measurement. |