Measuring three-dimensional (3D) geometrical parameters of weld pool surface is a key to developing the next-generation intelligent welding machines that can mimic a skilled human welder to control welding process through observation. It also provides reliable data to improve and validate welding process modeling. Although different techniques have been applied in the past few years, the highly dynamic and specular natures of the high temperature molten liquid surface complicate these approaches in addition to the well-known difficulties associated with the bright arc.

In this paper, a three-dimensional vision-based measurement system for gas tungsten arc (GTA) weld pool surface is proposed. In the system, the specular surface is taken as an advantage to overcome the difficulties caused by the bright arc through projecting a structured laser pattern onto the weld pool surface and imaging the resultant specular reflection on an image plane. It is easy to see the distortion of the reflected pattern caused by the deformed molten weld pool surface. This approach also allows the continuous reflection from the specular weld pool surface produced by a low-power illumination laser be used for high speed measurement of the weld pool surface.

To successfully derive the pool surface from the imaged specular reflection of the projected laser pattern, robust image processing algorithms, correspondence simulation method and iterative reconstruction algorithms have been developed. Using these algorithms, dynamic weld pool surfaces have been successfully measured/derived. The authors have also studied the fluctuations of the weld pool surface during gas tungsten arc welding (GTAW) process and conducted a systematic error analysis which can help further improve the measurement accuracy. Experimental results verified the effectiveness and accuracy of the proposed measurement system.