Strange Optics Ques: photon capture and reprod

[JS85]

Strange Optics Question: photon capture and reproduction

Imagine that you are sitting in your room staring out your window at a tree branch and a single leaf about to blow off. Suddenly someone comes up to you and says "You know that window, that's no ordinary window, it is actually a black sheet, with one side covered in photon capture devices and the other side covered in teeny tiny lasers. It only looks like a window because the properties and direction of each individual photon are measured on one side and the photon stream is reproduced on the other by tiny little lasers (like these http://gizmodo.com/5500375/the-worlds-smallest-laser).

Photoreceptors capture diffuse light, without a lens, you will just get a gray mess. However, lenses capture the properties of a light field at one specific focal point or range. But if you could capture the individual photons, you could duplicate a light field as it exists in the real world. So... is this theoretically possible? Could you have such a magic window?

 

[johng23]

I don't know, but it seems like a waste of money.

 

[DLuckyE]

Basically closest you can get is by just putting a camera on one side of the window and a monitor on the other side. Biggest problem being of course it'll only look like a window when viewed at the same angle as the camera is pointing.

Which is something I doubt you could get around, don't think it's possible to determine the direction of every photon that hits a certain location. And even harder would be to make a device that can send multiple photons in different directions with different properties from one location. Since a laser can only send them in one direction it won't be good enough.

 

[Andy Resnick]

Could you have such a magic window?

Like this?

http://www.thegreenhead.com/2004/05/japanese-scientist-invents-invisibility-cloak.php

 

[DLuckyE]

Like this?

http://www.thegreenhead.com/2004/05/japanese-scientist-invents-invisibility-cloak.php

I've seen this on tv, they make it sound complicated on some sites but basically all they're doing is the normal green screen routine with the difference being an extra camera added. If you're standing next to the guy it's still just a guy wearing a green coat.

 

[mrspeedybob]

Basically closest you can get is by just putting a camera on one side of the window and a monitor on the other side. Biggest problem being of course it'll only look like a window when viewed at the same angle as the camera is pointing.

Which is something I doubt you could get around, don't think it's possible to determine the direction of every photon that hits a certain location. And even harder would be to make a device that can send multiple photons in different directions with different properties from one location. Since a laser can only send them in one direction it won't be good enough.

You might not have to duplicate every photon, you are most interested in the ones that end up in someones eye. You would have to have multiple cameras on the outside of the window feeding data to a computer that would assemble a 3D image of what is happening outside. You would also have cameras on the inside of the window tracking the position of the observers eyes and pointing the lasers (at very low power of course) into the eyes of the observer reproducing the image that they would have seen had the window been glass. Additional lasers could mix and direct light towered the inside of the room to mimic the ambient light that should be shining in the window. I doubt a setup like this could make an absolutely perfect cloaking device, but it may get close enough that only the most astute observer would notice something off, and even then they may not be able to tell what it is.

 

[HallsofIvy]

You would still need some kind of lens because every point on a window receives light from every point outside. To be able to transmit, on the other side, exactly what you would see from the inside of a transparent window, you would need to have "cameras" that captured the light hitting the window only from the direction that would then continue to the eye of the observer.

Worse, different observers, at different points inside the window would be looking at it from different angles and would see different things through the window. If you have just one camera at each point on the window, only one person, at exactly the right angle from the window, would be able to see exactly what they would through a transparent window.

 

[mrspeedybob]

You would still need some kind of lens because every point on a window receives light from every point outside. To be able to transmit, on the other side, exactly what you would see from the inside of a transparent window, you would need to have "cameras" that captured the light hitting the window only from the direction that would then continue to the eye of the observer.

Worse, different observers, at different points inside the window would be looking at it from different angles and would see different things through the window. If you have just one camera at each point on the window, only one person, at exactly the right angle from the window, would be able to see exactly what they would through a transparent window.

I wasn't suggesting you would transmit exactly what the camera saw. You would use the cameras to construct a 3D model in the computer of what was happening outside. Then you would locate each eyeball in the room behind the "window" and calculate the image that particular eyeball would see. Having more then 2 eyes in the room would simply require more computing power and more lasers. Once these images are calculated you use the lasers to project images into each identified eyeball individually. Basically this is a no-optical-tricks-all-computer-driven approach.

 

[JS85]

So it seems that there is some degree of consensus that the most effective way of solving the problem is by using multiple lenses at various locations to estimate the "light field", using something like a plenoptic camera, (here: http://graphics.stanford.edu/papers/lfcamera/), and then reproducing the light field in another location... using lasers, overall, there seems to be a lack of response to this thread... I wrote the post in haste after my first post was deleted because of an auto-log out (they should save the form fields when they make you log back in), so I apologize if it is poorly described, aesthetically displeasing, or otherwise unappealing. Nevertheless, I appreciate your responses, keep em coming!

 

[JS85]

I actually had no knowledge of plenoptics or 4 or 5d representations of light http://en.wikipedia.org/wiki/4D_light_field#The_4D_light_field

before I made this post. But it seems like, for a lensless camera to work, it would have to be able to measure, or infer the 5D characteristics of each ray at each point, which could be done, in theory if you could pick out the photons, and represent it as a 5D vector (or higher if you are measuring polarity, wavelength etc).

 

[JS85]

Haha... Just one more thing, this video is beautiful... http://graphics.stanford.edu/papers/lfcamera/lfcamera.avi