BF2S Forums Is there such a program?

I lol every time that happens on CSI.

It works and you can do it. However, and that is a big HOWEVER, what the computer generates is complete guesswork and is in no way accurate.

There's been cases where enhanced photographs have been used as evidence during court proceedings, but that evidence has often been dismissed due to to inaccuracy of such a technique.

As such

The principal requirements to admit a photograph (digital or film-based) into evidence are relevance and authentication. Unless the photograph is admitted by the stipulation of both parties, the party attempting to admit the photograph into evidence must be prepared to offer testimony that the photograph is an accurate representation of the scene. This usually means someone must testify that the photograph accurately portrays the scene as viewed by that witness.

So if you have a blurry image of a killer, unless there's a witness to backup the computers complete guess work, it usually counts for nothing (not always, but in most cases).

GodFather wrote:

DefCon-17 wrote:

I lol every time that happens on CSI.

Its been on Numb3rs too... I mean mathematically it seems like a possibility, but then again my math education has only been to geometry thus far, so I dont get how it would work but I could see it

I think it's theoretically impossible.

Say a camera takes a max-res/quality picture of a city from a low-ish airplane and it is saved, you won't be able to see the details of the people walking below. Once it's saved and all that, how is any computer/program supposed to know what the details of the people look like?

You can't just take a pixel in a photo and split it into differently colored pixels to enhance detail.

GodFather wrote:

Okay, lets say this, you take a 900 by 900 picture into paint, and then take that grafitti tool and click it once with white, save it and run it through the program, the program takes the other pixels around it and guesses what colors they are and puts them in...

thats what im talking about, not making 80x80 800x800

I don't know about a program doing that, but I would be able to do that myself. Provide you don't hold the spray can button down for too long. lol

It's just basically looking at the pixels and adding them.

We thought you meant getting one pixel, and making it into 9 pixels, which is essentially what you do when you zoom in.

TheAussieReaper wrote:

GodFather wrote:

Okay, lets say this, you take a 900 by 900 picture into paint, and then take that grafitti tool and click it once with white, save it and run it through the program, the program takes the other pixels around it and guesses what colors they are and puts them in...

thats what im talking about, not making 80x80 800x800

I don't know about a program doing that, but I would be able to do that myself. Provide you don't hold the spray can button down for too long. lol

It's just basically looking at the pixels and adding them.

We thought you meant getting one pixel, and making it into 9 pixels, which is essentially what you do when you zoom in.

Fuck, I sort of lost my original train of thought

I mean clearing up an image, by adding more pixels, certainly not 9x pixels, maybe 2x, by looking at the whole picture and guessing what pixels lie inbetween


Can anyone phrase this better? Like someone who has seen it done (in the shows)

DefCon-17 wrote:

So you mean like the CSI episode where some guy took a picture facing away from where some person got hit by a bus, and they zoomed in on the photo subject's eye to see the reflection of someone pushing that person in front of the bus?

Yes, exactly like that, and like the episode where there was a refliction from the lamp, and they zoomed in and corrected the wavy distortion from the round lamp

No, there is no way to "zoom and enhance the image" like they do on CSI.

A computer can not guess, it computes.

Your other way of describing it is using the nearby pixels to guess what this one would be. Well, if X pixel is next to Y pixel and Y pixel is blue, what is X pixel? Red? Probably not, but if you're taking a picture of a sunset over the ocean, it's possible. But put in to words, it makes no sense. What if the computer makes X pixel blue based on the nearby pixel of that color? Then you've got a blue sunset, which also doesn't make sense.