This is (yet another) good example of a computational camera: where you can exploit megapixels, physics, and computation to build more capable cameras. There was a really good plenary talk by Prof. Nayar (Columbia) about other techniques a few years back:
My personal favorite is the "flexible depth of field," where you can actuate the imaging element during image acquisition and then use DSP techniques to create an image that is in focus across all depths simultaneously.
Also... if you'd like to get involved in these type of projects, a team from Stanford is working on the "FrankenCamera" -- a open-hardware and open-source effort:
http://www.hizook.com/blog/2009/06/26/computational-cameras-...
My personal favorite is the "flexible depth of field," where you can actuate the imaging element during image acquisition and then use DSP techniques to create an image that is in focus across all depths simultaneously.
Also... if you'd like to get involved in these type of projects, a team from Stanford is working on the "FrankenCamera" -- a open-hardware and open-source effort:
http://graphics.stanford.edu/papers/fcam/