Pictures in the clouds

Do you remember laying back in the long grass and gazing up at the weird and wonderful cloud formations? Back in the days before you had to work for a living, naturally. No time for that now. But it seems Ned Lecky still has time for a little cloud-gazing, and he describes what he sees in “Integration Corner: Cloud Computing” (Vision & Sensors magazine, March 1st, 2011.)

Some time back I wondered out loud if there could be some kind of cloud computing model that would advance machine vision, and Ned seems to be thinking along the same lines. He makes two points. First, that hardware/software compatibility issues are a big problem in vision and suck up much development time that should be put into meeting customer needs. Cloud computing, using the Google model as an example, shows how these can be eliminated. And second, he reasons that if we can play games with sophisticated graphics over the internet – and these are really just pixel –processing routines – then it should be possible to do the same with image processing.

I like the way he’s thinking, although I’m still not entirely sure how this vision translates into useable machine vision technology. One of the keys appears to be the use of HTML5, so perhaps the real takeaway from Ned’s article is that we all need to learn more about what this can do for us.
It’s this kind of breakout thinking that we need more of. Thanks Ned.