Font Size: a A A

The Weld Detection System Research And Software Design Based On CCD

Posted on:2018-05-26Degree:MasterType:Thesis
Country:ChinaCandidate:Y J ShengFull Text:PDF
GTID:2348330512473503Subject:Control Engineering
Abstract/Summary:PDF Full Text Request
Welding technology of automation and intelligent is to improve the welding quality,improve production efficiency and reduce manufacturing costs.Using robot welding has become the prime symbol for the automatic welding technology.Because of the vision sensor for non-contact measurement,large amount of information available has the advantages of strong sensitivity,high accuracy and strong anti-jamming capability.The welding control system based on visual processing has become the inevitable trend of development of welding robot.In order to realize weld automatic control,the weld visual detection technology is one of the core information technologies.At present,the intelligent automatic welding is not high,and application problems existed.Therefore the development of a high intelligence,strong anti-interference ability,fast processing speed,can reduce the error of weld visual inspection information system is very necessary.At the same time,the system suitable for high temperature,toxic,high dust and radioactive substances such as bad operation environment,or in some people cannot arrive within the scope of use,can work continuously and stable welding quality,can provide not only improves the work efficiency,reduce the labor intensity of workers,and reduces the production cost.In view of the complex welding environment,this paper designed a kind of automatic welding control system based on industrial CCD camera.In order to obtain accurate CCD image,the light and structure had been designed,to filter the reflective interference of workpiece,to improve the antiinterference ability.And the visual inspection system has been calibrated,including the camera and hand-eye calibration.By visual sensor in order to obtain the weld image can be extracted quickly and reliably weld feature information,through the in-depth study of image processing method,according to the characteristics of the visual sensor weldimage,using a one-dimensional discrete Lo G(Laplacian of Gaussian)filter,based on prior knowledge to improve the maximum variance threshold segmentation and extraction methods of local information,only to get the clear fast weld image.After get weld image processing algorithm skeleton,extract the weld characteristics of reliable information.To realize automatic and intelligent welding control system,the control interface has been designed,using Open CV,Halcon,MFC compiled a set of mixed weld information detection system,the system is of friendly interface.Experiments show that the detecting system in high accuracy(2mm),fast processing speed(0.8s).
Keywords/Search Tags:Double location, Argon arc welding, Vision sensors, Weld extraction
PDF Full Text Request
Related items