Extract the aluminum alloy MIG welding pool edge with Otsu threshold selection based on the genetic algorithmic

摘要

It is one of the important methods to detect the image edge of the molten bath for gain the characteristic information including the maximum width of molten bath, the half length of the molten bath and the rear angle of the molten bath etc. It is one of the important methods to achieve automatic control for welding process that the visual sensing technology is used to gain the pictorial information of molten bath and make processing. This paper introduced the structure of experimental system and algorithm, uses the Otsu to realize the detection of aluminum alloy MIG welding molten bath edge, and compared with the Sobel operator and the Canny operator molten bath edge detection result, the experimental result indicated that it is completely feasible using genetic algorithmic to realize the detection of welding pool edge based on threshold value selection, and the molten bath edge information obtained is smooth, continual, without the break point and the noise.