I'd break the problem down into parts:

1.  Use one of the computer vision libraries (e.g. openCV) to figure out how to track a person.  You could do face tracking - though bear in mind that they'd need to keep facing the camera as they move - or difference tracking against a stored background.  The libraries should include plenty of examples ;)

2.  Figure out how to control the servo from Processing using a numeric input (again there should be examples on the Arduino site).  Also build in some form of calibration - though this might be as simple as multiplying by a variable that you can adjust manually; though an interface would help speed up calibration in the next step.

3.  Combine the two sketches.  Assuming you want to track the person with the lamp, I suspect you'll have to overcome a couple of issues: firstly face recognition and blob tracking can be a little erratic - e.g. jumping from one individual to another; so you might want to build in some sort of selection process or smoothing.  Secondly the angle you need to point the lamp at will depend to some extent on the distance the user is from the lamp.  Here face recognition might have an advantage as you should be able to use the 'size' of the face to determine a rough distance.

That's all hypothetical of course.  This isn't something I've tried, though I have certainly considered it.  It's not especially trivial - though judicious use of existing libraries will certainly help a lot - and might be quite an undertaking if you have no coding experience.  It looks like I may have a boring Christmas holiday ahead and might be looking for an interesting project to play with.  If I do decide on something similar we can compare notes ;)