Interesting. I would like to see that done. I don’t think it would be as trivial as simply reverse engineering a gaussian blur and doing the obverse of the setting to put it back.
Once the data is destroyed (as it would be in a jpg once you apply the blur and save the image) the data is gone.
The analogy would be a picture that was not exposed correctly. There is a limit to the amount of manipulation that can be done on an image if the data simply isn’t there. If someone is wearing a black shirt and the basic image doesn’t have enough shades of gray to distinguish detail, then you will never get the detail out because it isn’t there to begin with.
I went out and looked around, and I did find a white paper that explained how it might be done mathematically, so I follow the rule of thumb: If someone can think of it, someone can probably do it.
I suppose it would depend on the type of filter that is used. If there is one that uses a more random algorithm to produce an effect, then it probably becomes a lot more difficult to reverse engineer.
I will keep that in mind for the future. Thanks for the tip.
Please do not underestimate the ability to increase resolution or restore “lost” information. It’s not as hard as you think.