An image process way of globular transfer in MIG welding

This paper carried out an image process way about feature extraction of droplet transfer. The adopted materials are that welding wire diameter is 1.2mm and sheet steel thickness is 3.0mm. In order to obtain high quality images of dynamic droplet, a special filtering lens for image filtering with CCD was designed and then an algorithm of image was developed. Under MATLAB software platform, the algorithm compared median filtering, images sharpening, morphological operation, threshold segmentation, edge detection and so on. Because of images´ magnification scale coefficients of original image, this paper calculated the coefficient according to characteristic parameters such as wire´s diameter. Results proved that the image processing method could deal with the obtained droplet images and get their critical features effectively.