Research On Automatic Positioning Technology Of Electrical Connector Glue Injection Based On Machine Visio

At present,the gluing of electrical connectors in the process of making cluster cables is still mainly manual,but with the demand of cable automation production,it has become a need to study and design an automatic equipment to replace manual gluing.The key of automatic gluing equipment lies in the identification and positioning of the gluing mould.Therefore,based on machine vision technology and binocular vision principle,this paper studies the identification and positioning of the gluing mould of electric connector.The main work contents are as follows:First,the camera calibration is realized.According to the basic principle of Zhang Zhengyou’s camera calibration method and camera iterative calibration method,an automatic iterative calibration algorithm based on ring feature is proposed,which realizes camera calibration through checkerboard calibration image with ring.Experiments show that the calibration results of the camera calibration method proposed in this paper are similar to those of the Matlab camera calibration toolbox,and can significantly reduce the reprojection errors of the calibration results.Secondly,the mold identification and location were realized based on feature point template matching.In order to solve the problem of high mismatching rate of feature points extracted by feature point detection algorithm,feature point violence matching method and RANSAC algorithm were introduced,and SURF algorithm and SIFT algorithm were combined to conduct experiments.The experimental results show that the violence matching method and RANSAC algorithm can significantly reduce the mismatching rate of feature points.After that,the matching point pairs of feature points based on SURF algorithm,violent matching method and RANSAC algorithm were used to realize the identification and positioning of the mold.The coordinates of the center point of the mold were determined,and then the center point of the mold was located by SGBM distance detection algorithm.Finally,based on QT5 design and write the mold visual positioning interface program.The experimental results show that the system can realize the identification,positioning and ranging of the rubber injection mold reliably.The depth measurement accuracy is higher within1000 mm,and the error is less than 1%.