It's very rare that I see something in Image Processing that makes me go 'wow'. The fact is that a lot of the current state-of-the-art in the field is becoming increasingly limited in its scope, and many good research groups are either wallowing in their own success or pursuing ideas that are more and more mathematically complex yet lack any real purpose.
This video on youtube though is brilliant. From a technical point of view, the work is quite basic. But from a 'bugger me that's such a cool idea' point of view, it is superb. It just goes to show that the best ideas are not necessarily the most complex or technical ones.
And I can't help asking myself why I didn't think of that for my PhD?! I'd have made millions!