News for hackers. Stuff that matters.
Re:Mathematicians Encrypt Images Using Mathematics of Sudoku by aneth4 on Friday July 27, 2012 @08:51AM (#4301098)
Since a sudoku puzzle implies it's solution, it is a highly compress(able) representation of the solution. I wonder if it might be possible to leverage this sort of rule based matrix to improve compression algorithms, or if compression algorithms are already more optimal than this.
Re:Mathematicians Encrypt Images Using Mathematics of Sudoku by breckinloggins on Friday July 27, 2012 @09:25AM (#4301240)
That was my thought as well (see my comment elsewhere in the thread). I wonder if this could be used as some kind of fixed sparsity encoding?
Re:Mathematicians Encrypt Images Using Mathematics of Sudoku by switch33 on Friday July 27, 2012 @09:41AM (#4301298)
I'm no expert but just an thought:

A sodoku puzzle(requiring only 17 digits for an unique solution) which has an single solution isn't exactly useful for compression.

How I am thinking about this? The article states:

"And since for a 256 x 256 image, there are at least 256!=2^1684 possible Sudoku matrices, it's not easy for an adversary to hit on the solution by accident or even by brute force. "

Compression would make sense if the number of puzzles was something like 1684, but it is here 2^1684!! Which is quite a lot and to be decoded would be even worse than the image itself.

The sodoku puzzle's compression would not be of any practical use not only because they give the data more problems to account for but also because the data would need to be interpreted.

Also anything that uses encryption generally does the complete opposite of compression.

Re:Mathematicians Encrypt Images Using Mathematics of Sudoku by aneth4 on Sunday July 29, 2012 @01:27PM (#4309065)
For sure it increases the computational power required to decompress, but that does not preclude its usefulness. Sometimes computation is cheap and bandwidth is expensive, as perhaps when sending messages to distant stars.