Vision-based identification model and extracting algorithm of wire extension in aluminum MIG alloy welding process

A vision sensing system and corresponding imageprocessing algorithm for extracting the information of wire extension during aluminum MIG (metal inert-gas) welding process have been presented.The images of welding area are real-time captured by CCD camera.According the character of the wire in the image, its ordinate can be located by image processing algorithm.After image clipping, shifting and superposing, the difference between wire and background in the image will be enhanced, and then the information of wire extension can be extracted accurately.Experiments show that this method is effective and feasible.Based on the step response experimentation, the influence of wire speed on wire extension is identified and the mathematic model has been set.It provides the theoretical basis to realize the wire extension sensing and control in aluminum MIG welding process.