Effective method for monitoring the welding process is essential to improve welding quality [4-6]. For instance, seam tracking system LPF from Precitec, welding monitor from Prometec, and RoboFind from Servo Robot have been

Process monitoring is important for the laser welding system, especially for the weld seam deviation monitoring. A 10kW high power fiber laser butt joint welding of type 304 stainless steel was performed. By combing two different band pass filters, high speed photography was employed to capture the near-infrared images of molten pools. Image difference algorithm was used to obtain the weld seam position. By accumulating numbers of near-infrared image, the weld deviation could be monitored. Three different kinds of welding trajectory were employed in testing the effectiveness of weld seam deviation monitoring by using image difference algorithm. The experimental results showed that the accumulation of difference images could be used in laser welding seam deviation monitoring. The maxima gray-value of accumulated image indicated the weld seam position while the laser beam focus position in the image was constant.