What math do they use and how does it work?

[nec209]

I have 2 quetions what math do they use to put a hidden picture or information in a picture like a Jpg ? And how does this work?

 

And what math or software do they use to make picture bigger or make it clear.

 

That say a security camera like a robbery and they use cheap camera and low quality .What math would they use to make it bigger and clear .

 

 

Low quality and small picture.

http://www.blogcdn.com/www.joystiq.com/media/2008/12/gamestoprobbery.jpg

 

I thought the video enhancing department of the police station use adobe photoshop ? Some one was saying there is no limit to video enhancing still image or moving image it is just more math.

 

How does this work and how long does this take? Can person at home with no training or little training do it?

[Cap'n Refsmmat]

There really is a limit. You can't introduce detail that simply wasn't in the original image. Photoshop could sharpen that image up a bit, but not make a nice image of the attacker's face.

[ydoaPs]

Well i'll be damned "The face of Mars" is a Lion.

 

 

 

 

 

Source

 

What Cap'n said...

Edited May 25, 2009 by ydoaPs

[nec209]

I was just reading today that they use the same math to make pictures of space look better has the same math to make a picture look better.

 

Saying it just more math .If one take small grainy picture you can make it bigger and clear it just very advance math.

 

Not sure what math they are talking about.

[insane_alien]

you can't indefinitely clear up a picture. you can do stuff like increasing contrast or brightness but if the information isn't there then you can't do anything with it.

 

it someones face is only on one pixel then you cannot 'enhance' the picture to get a recognizable face. at best you might get a face coloured blob that's still roughly square.

 

the enhancements don't actually add any details, they just make it easier to spot details already present in the picture that the human eye is bad at picking up on. light if it is a low contrast night scene, you may be able to pick out people and objects if you increase the contrast and brighten the picture.

 

EDIT: here is the main culprit for why people think images can be infinitely enhanced

 

http://www.makemymood.com/2009/10/21/csi-zoom-story/

[the tree]

if the information isn't there then you can't do anything with it.

This.

 

But you can use fairly simple maths to highlight various details, such as edge detection to spot major discontinuities and noise reduction to disregard less significant ones.

[uncool]

It is possible to figure out more about a face than a picture has, but that is more from knowing about faces than using math. That is, you cannot simply "enhance" a jpeg past its original resolution using math, but if you have a good picture, you can rotate it so that it's in a certain direction, and use certain facts about faces to figure out a little more about this particular face. This is extremely difficult, but possible.

=Uncool-