| Plasma Arc Welding technology, using rapid plasma arc as a source of heat melting and using NC machine's control technology to follow the direct path of manufacture process, is a new kind of technology for rapidly forming metal parts or prototyping manufactur.At present, this kind of technology has unique advantages in the field of the rapid manufacturing of functionally gradient materials and high temperature carbide parts. Because it provides a good solution to the rapid manufacture of high temperature alloy parts , more and more reseach plans were carried on how to use of this technology as the same time it becomes the most popular and most important development direction of Rapid Prototyping Mafuture technology. Welding quality control has always been the most difficult problem in welding field, and also the most problematic area in the study. That How to conrtol the plasma forming welding quality to ensure uniformity of weld is the focus of this paper. weld morphology detection technology is the most critical technology in quality control application. Throught detecting the weld beam geometry shape and sending the width deviation of weld bem directly and effectively to the controller, to work out a real-time feedback and control to make online adjustment of process parameters to reach the process optimization, for forming components to provide the ultimate precision and quality assurance. In this paper, the vison sensor applications was studied, and a a passsive vision detection technology was applied in plasma processing successfully.Plasma welding is a complexity process and the welding environment is extremely worse. In particular the arc and splash disturbance cause a large amount of noise. These interference cause trouble and great difficulties to Image process In this paper, using a narrow-band filter method of visible light, UV filter and the rest lights reached the CCD were used as light source, then we can use a CCD cammera to grab a clear image of the weld beam. Image processing software can obtain image data. From Adlink's RTV-24 frame grabber And in memory to work out many tasks. Such as image preprocessing, feature recognition, the width of deviation, the image data output. The key factor to work a good job lies in the image detection image processing algorithms. During the weld vision detection, edge detection is the premise of the weld size calculation.Under the Windows XP system platform ,we use VC + +6.0 programme software finished weld dectect programme modules division and code-preparation. We realize the image acquisition function with the library functions provided by Image acquisition card. In this paper, we firstly applied Canny algorithm Combined with image windowed processing method to finish task of weld edge detection successfully. The first step is to get a smooth image using Gaussian filter; the second step is to calculate the gradient through a first-order partial derivatives of the finite difference; The third step is to work out the non-maximum value suppression of Gradient amplitude; The last step is to make use of dual-threshold algorithm for edge detection and connection. Our programme can finish the manual and automatic measurements tasks through user-specified width measurement of the direction of the width of a weld...
|
PDF Full Text Request