Research Of Stud Workpiece Size Detection Algorithm Based On Machine Vision

In recent years,machine vision detection technology has been rapidly developed due to its advantages such as high detection accuracy,fast speed and non-contact,and it is also one of the major development trends of workpiece size detection.In this paper,based on the problem of stud workpiece size detection,the automatic detection system and the detection algorithm are studied and tested,which using the related technologies and methods of machine vision.The paper mainly completes the following work:1.From the analysis of technical requirements,an automatic detection system hardware platform consisting of CCD camera,telecentric lens,parallel light source and industrial control computer is built.2.An algorithm for dimension detection of stud workpiece is presented based on machine vision,which is to perform grayscale processing,image denoising,edge detection,sub-pixel precise positioning and contour segmentation for the collected images.After the image grayscale processing,based on the characteristics of the stud workpiece,the pseudo-median bilateral filtering method is proposed for image denoising,and the optimal Canny operator is adopted for image edge detection.On this basis,the gradient interpolation method is used to analyse the image edge pixel level and accurate positioning.Then,the Ramer algorithm is used for contour segmentation to extract all the feature points in the image,which provides a good basis for calculating geometric parameters such as pitch,tooth angle and tooth height.3.The calibration of detection system is done precisely and calculate the calibration coefficient K.According to the measuring principle of the parameters such as pitch,tooth angle and tooth height,and the detected pixel values,the actual dimension values of each parameter of the stud workpiece are obtained respectively,and the error analysis is performed.The experimental results show that the detection accuracy of the stud workpiece size can reach 0.08 mm,and the accurate detection rate can reach more than 98%,that this algorithm can meet the actual demand.