Dennou Coil glasses
So, I'm watching Dennou Coil, and these glasses are pretty awesome.
Most of it is pretty simple stuff, in real life we wouldn't be able to have the glasses as 'stand alone' so you would need to have a processor in a backpack or pocket that is either wired or wirelessly linked, it wouldn't need to have any extreme power. Then that would hook up to the Internet, which I guess would need some kind of system or directory to find virtual topology.
I imagine an initial system using nodes rather than a city or country database; for example a shop could buy a 'node' which would be a small wireless router sized device, able to broadcast the node information, such as the 3d model to use and it's location. Properties of the node may be sent too, such as 'isTouchable' flag (for example for a virtual beachball), while most of the processing is done by the node (for example for fluid simulation). I guess the only thing about this solution would be determining node location. The only way I can think of to generate an accurate location data means needing direct line of sight, as I'm going to say later, or with some kind of wireless triangulation, which is not quite as precise as I would like.
The other idea is to use 'tags' as is used for current augmented reality, which looks like a QR code looking thing, and when an AR device sees the tag it can interpret them as a texture or object or whatnot. I have seen some really interesting things about using multiple tags (for example 4 in a square), so that a device need only see one of them to interpret the object. Anyway, obviously this would have a different use to a 'node', as the tag would have a web address that the server responds with the object to put in to the space, I imagine these to be instantiated rather than constants.
The nodes solution requires a way to track user position and facing, the tags need some kind of camera. The camera would be easy to do really (that is pretty standard now days).
I wonder what kind of technology would be best to reproduce the effect? The two most obvious ones to me are projection and transparent LCD. Transparent LCD would be great, but currently it is what, 60% transparent? Projection would be cool, but the problem is resolution and the size, plus it may be hard to get it in a position where hair wouldn't block it. LCD would be best if they would hurry up with that technology... Also, the thing that makes this type of technology easier is that eye position doesn't matter, since they are glasses.
The virtual object interaction thing is kinda tough. In general it would be easy to detect hand positions and whatnot. The real problem is just doing it with the glasses. Obviously there would need to be a z-map generated somehow, but it would also need to be able to sense finger locations, and usually your hands will block your fingertips from your eyes when you type, so it would be inaccurate. The best part of the technology just happens to be the hardest to reproduce I guess.
Anyway, I want some.
Most of it is pretty simple stuff, in real life we wouldn't be able to have the glasses as 'stand alone' so you would need to have a processor in a backpack or pocket that is either wired or wirelessly linked, it wouldn't need to have any extreme power. Then that would hook up to the Internet, which I guess would need some kind of system or directory to find virtual topology.
I imagine an initial system using nodes rather than a city or country database; for example a shop could buy a 'node' which would be a small wireless router sized device, able to broadcast the node information, such as the 3d model to use and it's location. Properties of the node may be sent too, such as 'isTouchable' flag (for example for a virtual beachball), while most of the processing is done by the node (for example for fluid simulation). I guess the only thing about this solution would be determining node location. The only way I can think of to generate an accurate location data means needing direct line of sight, as I'm going to say later, or with some kind of wireless triangulation, which is not quite as precise as I would like.
The other idea is to use 'tags' as is used for current augmented reality, which looks like a QR code looking thing, and when an AR device sees the tag it can interpret them as a texture or object or whatnot. I have seen some really interesting things about using multiple tags (for example 4 in a square), so that a device need only see one of them to interpret the object. Anyway, obviously this would have a different use to a 'node', as the tag would have a web address that the server responds with the object to put in to the space, I imagine these to be instantiated rather than constants.
The nodes solution requires a way to track user position and facing, the tags need some kind of camera. The camera would be easy to do really (that is pretty standard now days).
I wonder what kind of technology would be best to reproduce the effect? The two most obvious ones to me are projection and transparent LCD. Transparent LCD would be great, but currently it is what, 60% transparent? Projection would be cool, but the problem is resolution and the size, plus it may be hard to get it in a position where hair wouldn't block it. LCD would be best if they would hurry up with that technology... Also, the thing that makes this type of technology easier is that eye position doesn't matter, since they are glasses.
The virtual object interaction thing is kinda tough. In general it would be easy to detect hand positions and whatnot. The real problem is just doing it with the glasses. Obviously there would need to be a z-map generated somehow, but it would also need to be able to sense finger locations, and usually your hands will block your fingertips from your eyes when you type, so it would be inaccurate. The best part of the technology just happens to be the hardest to reproduce I guess.
Anyway, I want some.
