In the laser enhanced gas metal arc welding (GMAW) process developed recently, droplets of melted metal can be detached from the wire under relatively low currents with the assistance of an auxiliary force provided by a laser. The stability and the arc and the quality of the resultant welds are improved. To compete with the gas tungsten arc welding (GTAW) of the much lower productivity in joining precision, the size of the droplet can be pre-defined and be controlled to meet the requirements from different applications. For this purpose, image processing algorithms are developed to measure the size of a growing droplet during the laser enhanced GMAW process. The relatively low contrast, strong illumination and reflection caused by the laser, and strong radiation from the arc make an automatic processing of the image challenging. Images are analyzed to understand its characteristics and design the image processing and recognition algorithms accordingly. In particular, a model based method is used to filter out non-droplet edge points and a second order equation in the polar coordinate system is introduced to model the droplet. Experimental results verified the effectiveness of the developed algorithms.