Abstract: By capturing the welding arc images at different time with the help of high-speed camera, and then recognizing their edges and extracting their characteristic parameters via images processing, the quantitative evaluation on the stability of arc burning and welding process could be achieved. On the basis of analyzing the grey characteristics of arc images, two simplified algorithms achieving the edge recognition of arc image were proposed based on the traditional and morphological image processing respectively. And the merits of Sobel edge detection algorithm in traditional image processing and the edge detection method in morphological image processing were compared from the two aspects of the program run time and the edge detection precision of arc image. According to the comparison result, the simplified algorithm of traditional image processing was chosen and applied into the construction of image batch processing algorithm. The performance of the program was tested by adding random gaussian noise to the original arc image and accurately identifying the edge of TIG-MIG hybrid arc image. The results showed that the image batch processing algorithm had good anti-interference ability and applicability.