Nowadays with digital photography and imaging consumers and manufacturers are in a mega-pixel craze. The next best thing always has more resolution crammed into the same space for a higher quality image. And then the bells and whistles kick in with the ability to record audio/video, lcd touch screens, a flash, zoom capability and the works. However, in most senses, the technology is the same. It's light that gets captured as a certain color within a certain position in the view of the lens, which then gets converted to a pixel of an image that gets converted into some sort of raster technology like the JPEG.
But what if I wanted to say, "locate the red VW Jetta"? The camera has no understanding of what that is. The camera doesn't know Johnathan from Natalia, even though we do. I think we're moving into a time where the consumer will want that. Just as they are able voice activate their Ford Fusion and tell it what song to play, or to check their Facebook page, I don't think that wanting the camera to be able to distinguish one object from another and then appropriately capture it is far behind.
How do we do that? How do we develop an imaging system that works similar to the human brain in a way such that it can distinguish the difference between two people after being given the definition of them? Or between two cars, or even just two colors to start. I say "photograph the blue ball" and it focuses heavily on that in the image and puts the red ball out of focus.
Is that a hardware issue? Do we have to develop an entirely new way of capturing images that isn't based on pixels and raster technology? Is it just a software issue and we need to develop better software to reach that type of interaction with our imaging equipment?
The best example I can think of visually is Iron Man 2 -- the inside interface within the suit helmet. I know it's all CGI, but the way the software was able to scan the screen and distinguish people from objects, weapons from non-weapons, and threats from non threats is the type of ability I'm speaking of.
But what if I wanted to say, "locate the red VW Jetta"? The camera has no understanding of what that is. The camera doesn't know Johnathan from Natalia, even though we do. I think we're moving into a time where the consumer will want that. Just as they are able voice activate their Ford Fusion and tell it what song to play, or to check their Facebook page, I don't think that wanting the camera to be able to distinguish one object from another and then appropriately capture it is far behind.
How do we do that? How do we develop an imaging system that works similar to the human brain in a way such that it can distinguish the difference between two people after being given the definition of them? Or between two cars, or even just two colors to start. I say "photograph the blue ball" and it focuses heavily on that in the image and puts the red ball out of focus.
Is that a hardware issue? Do we have to develop an entirely new way of capturing images that isn't based on pixels and raster technology? Is it just a software issue and we need to develop better software to reach that type of interaction with our imaging equipment?
The best example I can think of visually is Iron Man 2 -- the inside interface within the suit helmet. I know it's all CGI, but the way the software was able to scan the screen and distinguish people from objects, weapons from non-weapons, and threats from non threats is the type of ability I'm speaking of.
I'm just sort of thinking out lout here...
Any thoughts?