I think it's interesting, but such an approach is still far to limited.
It's the same setup that we had for touch interaction.
People have been building and showing off multitouch interfaces since 1982 at places like TED. But only in the 3 years have they become ubiquitous (or at least
widespread), because companies have started to use them in ways that are fundamentally superior to our old way of doing things. Well that and the cost of
implementing them fell enough to make it commercially viable.
I can see three major problems that they need to overcome:
1) Camera needs to be mounted closer to eye level, otherwise you will always be dealing with parallax, and correcting for that is not cheap computationally.
2) We need better ocr, much better. Or we have to start to embed matrix-codes (2D bar codes) in a whole lot more places. The demonstration of grabbing bits of
data from different documents made me wince.
3) Batteries! Projectors suck power like it's going out of style, and our battery tech is only increasing linearly (if that) while our demands for portable
power are exponential.
These are not simple things to overcome, but they can be done. I just don't expect to see anything like this hitting the market for another 10 years.
-Terry
-----
"so listen up boy, or pornography starring your mother will be the second worst thing to happen to you today"
TF2: Spy
It's the same setup that we had for touch interaction.
People have been building and showing off multitouch interfaces since 1982 at places like TED. But only in the 3 years have they become ubiquitous (or at least
widespread), because companies have started to use them in ways that are fundamentally superior to our old way of doing things. Well that and the cost of
implementing them fell enough to make it commercially viable.
I can see three major problems that they need to overcome:
1) Camera needs to be mounted closer to eye level, otherwise you will always be dealing with parallax, and correcting for that is not cheap computationally.
2) We need better ocr, much better. Or we have to start to embed matrix-codes (2D bar codes) in a whole lot more places. The demonstration of grabbing bits of
data from different documents made me wince.
3) Batteries! Projectors suck power like it's going out of style, and our battery tech is only increasing linearly (if that) while our demands for portable
power are exponential.
These are not simple things to overcome, but they can be done. I just don't expect to see anything like this hitting the market for another 10 years.
-Terry
-----
"so listen up boy, or pornography starring your mother will be the second worst thing to happen to you today"
TF2: Spy