Method And Application Research Of Industrial Inspection Based On Adaptive Filtering

With the development of the times,the field of industrial vision inspection calls for higher and higher requirements for inspection accuracy.An adaptive filtering method is proposed to improve the accuracy and stability of defect inspection and speed measurement through adaptive filtering algorithms,which is applied to the surface defect inspection system and welding line speed measurement of automotive injection molding screws,and its reliability is verified through experiments.The specific research work of this thesis is as follows:(1)Aiming at the problem of low efficiency and poor reliability of automotive injectionmolded threaded parts by manual detection,a reliable detection method based on machine vision is proposed and an improved edge segmentation algorithm is proposed to improve the accuracy and stability of detection.Firstly,The canny operator is improved by adaptive filtering method and different enhanced image features are acquired combined with watershed algorithm.Then the support vector machine(SVM)is trained based on the image features of three typical defects such as incomplete filling,burr and encapsulation.This classifier can be used to detect the defection of the threaded parts.Finally,the experimental test platform is constructed with the UR5(Universal Robots5)robotic arm,and the automatic sorting of the threaded parts and the elimination of defective product can be realized.The experiment results show that this method can effectively detect the typical defects of threaded parts and improve the stability and accuracy of the system.Besides,the automatic detection and elimination of defective products can be realized with the robot arms.(2)In the production process of automatic welding,the welding speed is an important guarantee for welding line quality.In order to improve the accuracy and reliability of welding,a welding line speed measurement method based on adaptive camera calibration is proposed and an automatic measurement system is built.Firstly,accurate image features are obtained by combining adaptive filtering method with feature point extraction algorithm,and the image distortion correction is calculated to obtain the internal and external parameters of the camera.Then,the homography is used to get the specific pose of the calibration board and the welding line speed is calculated.Finally,an experimental platform is built for comparative analysis of experiments.The experimental results show that measuring the welding line speed with the method of adaptivecamera calibration is simple,rapid and high-precision,which helps to improve the detection efficiency and the degree of production automation.