The Motion > Stabilization 1 point effect

This effect works as follows :
For each image from the sequence, you have to define an Origin point and a Destination point. The effect will make a panning of the image in order to bring the Origin point onto the Destination point, which will stabilize the sequence.
Now let's take an example.

Let's take the background below : the cameraman has shaken a little bit while shooting the sequence. We can define one of the yellow points below as the Origin point, and estimate that this it should be at the position of the Destination point (at the extremity of the yellow arrow) if the camera had not shaken. As soon as these points are defined, the Stabilization effect will do a panning that will bring the Origin point onto the Destination point. The footage will not shake anymore.

Origin pointHorizontal and Vertical camera movementsDestination point

You can argue that each positions of the Origin and of the Destination points are different for each image, which make the process tiresome. That's why, at this stage, using a path for the Origin and/or Destination points can speed up the process, especially paths created with the Pixel Tracker...

As soon as the effect is applied and that the image is stabilized, you would have to crop it to remove the white borders due to possible panning.
You can also use the Mode button to remove the borders by filling the image.