Radon/Abel transforms

[r08n]

Dear all,

I'm looking for a way to reconstruct axially-symmetric images from their
projections. I have a couple of n00b questions:

Is there any function in Octave to do the inverse Abel transform (a.k.a.
Abel inversion/reconstruction)? If not, is it sufficient to call the
'iradon' function with an array of identical projections (assuming axial
symmetry) as the argument? This looks a rather clumsy but simple way to
achieve the Abel inversion... or do I miss something?

Another question is how to interpret the Radon projections returned by the
'radon' function: if the original image size is N*N, the  "longest"
projection is the one seen from the diagonal direction, so its size (array
length) must be sqrt(2*N^2); for an image with size 256*256, this would
amount to 362.04, and the actual size of projection is 367; likewise, for
other sizes. Is there some padding in the returned values?




--
View this message in context: 
http://octave.1599824.n4.nabble.com/Radon-Abel-transforms-tp4658783.html
Sent from the Octave - General mailing list archive at Nabble.com.