Although it’s said that most of Grey-Level Co-occurrence Matrix (GLCM) features range less than 1, I am wondering why SNAP produce the high values E.G., for Dissimilarity, entropy,…
Is that statistically wrong to report sth greater than 1?
On Sigma0 (linear). I am fine with the high values, but statistically, dissimilarity feature ranges between 0 and 1. But SNAP gives sth up to 35 even sometimes.
I think that 32 bit level defined might cause this. But not yet sure about. Just looking for a way to justify the SNAP range.