Slide 7 of 42
(1) is the ìsurface signatureî we have been using so far. It is simply the probability distribution of the segment lengths, found by normalizing the histogram of binned counts of segment lengths. Some auxiliary data is generated as well, which I wonít be
The 2-D histograms (2) which combine length and reflection-angle data look interesting, but in initial tests do not seem to contain any more information than the 1D histos.