Design Of Positioning And Identification Vision System Software For Mixed Flow Spraying Lines

The spraying operation adopts the robot teaching spraying and manual manual spraying operation,the production efficiency is not high,the skills required by the workers are high,and the health is harmful.Some companies use the mixed-flow spraying process to improve the production efficiency,resulting in a lot of paint waste.Increased production costs.In order to solve the above problems,the subject is based on a robot teaching spray line of a company in Foshan,using visual technology to solve the above problems.The research carried out in each chapter of this paper includes the following:(1)In order to solve the noise problem in complex spraying environment,the appropriate algorithm is selected as the implementation method of the functional module by comparing the experimental results of the filtering denoising algorithm.The camera calibration adopts the plane calibration method to establish the camera model to solve the camera internal parameters and external parameters.The internal and external parameters of the camera are used to realize the image distortion correction based on the distortion correction algorithm.The perspective transformation method is used to realize the image projection transformation correction.(2)In this paper,the camera plane calibration method is used to solve the single-image relationship between the camera coordinate system and the world coordinate system.The DH parameter method is used to derive the single-image relationship matrix between the robot coordinate system and the world coordinate system.Based on camera calibration,robot coordinate system and world coordinate system calibration,a new hand-eye calibration method is derived.The calibration error of the camera calibration and hand-eye calibration is verified by the parameter verification method.The method is simple,easy to understand,and meets the requirements of the project.(3)This paper writes the program for the commonly used edge detection algorithm,and compares the implementation effects of various detection algorithms;the edge segmentation method is used to realize edge segmentation.By implementing the above various methods,the processing results are not ideal,so the characteristics of the various methods are combined to achieve the effect,and an improved edge segmentation algorithm is proposed.The method can be better adapted to the generality of the spraying scene and the processing efficiency is higher.(4)Classification identification function is the key to solve the automatic identification of mixed flow products in the production line.Based on the comparison of template matching and support vector machine(SVM)classification and recognition,SVM classification has lower false positive rate and higher efficiency.The positioning function enables the robot to automatically detect its pose information based on the product type.The positioning function of the system software is realized by matching positioning and blob analysis and positioning methods.The visual system's classification and recognition of products and the realization of positioning functions are the premise and core links for automatic and intelligent spraying.(5)By implementing various functional modules of the system software,in order to verify the overall implementation effect,the multi-threaded system software program is written based on the VS development platform and the halcon image processing library,and the system software interface is designed;the three product testing system software programs are used to verify the system.Software holistic features achieve results.