[getsaf]

I can't help but think that these things cannot be too accurate. Even in the cheese example given in the article, the outer layer of many cheeses are slightly (some more than others) different than the inside of the cheese block. So how accurate could the reader be for giving contents of the entire block given a scan of a small surface area.

This gets even worse when you scan say, the outer shell of a ravioli. The reader would see pasta which is great but how does the reader know whether it's just pasta or pasta stuffed with lard or spinach?

Thoughts?