To　understand,　analyze,　and　control　the　metal　transfer　process　of　double-electrode　gas　metal　arc　welding　(DE-GMAW),　the　droplet　should　be　monitored　in　real　time.　Due　to　the　fast　development　and　transfer　of　the　droplet,　the　monitoring　speed　becomes　the　key.　A　tracking　method　that　takes　advantage　of　results　from　previous　images　to　speed　the　processing　is　advantageous.　In　this　paper,　Kalman　Filter　tracking　and　Least　Square　Match　tracking　algorithms　are　studied.　Finally,　a　real　time　template　updating　and　LSM　tracking　method　is　proposed　to　track　the　droplet　effectively.　Experimental　results　verified　its　tracking　accuracy.