[3dem] Total variation in ET reconstruction(?)

Johannes Haataja johannes.haataja at aalto.fi
Thu Sep 1 03:38:36 PDT 2016


Hello,
	Yes - more or less, TV assumes that structure is composed of regions of
constant density/minimizes contrast variations between neighboring
voxels. It will lead to blocky/staircase like reconstructions depending
on the chosen strength of TV regularization. Usually the biggest
advantage is edge-preservation in reconstructions (nice for interfaces),
in comparison to, let's say, smoothness reconstruction were edges are
usually heavily blurred. Whether it's appropriate for protein and
cellular structures, I don't know, but I work mainly with polymer and
inorganic samples so it's not my primary concern...

 - J. Haataja

On to, 2016-09-01 at 10:10 +0000, Henning Stahlberg wrote:
> Hi,
> Isn't that scheme assuming that within the structure there is no internal contrast variation?
> If so, that can be used to phase X-ray tomography data of plastic objects, but it would not be a fitting assumption for protein or cellular structures. 
> Henning. 
> 
> Henning Stahlberg
> HStahlberg at me.com
> http://c-cina.org
> Tel.: +41-79-502 75 22
> (sent from iPhone)
> 
> > On Sep 1, 2016, at 12:05, Johannes Haataja <johannes.haataja at aalto.fi> wrote:
> > 
> > Dear all,
> >    does anyone know if any of the freely available ET software packages
> > include total variation (TV)/compressive sensing reconstruction scheme?
> > I would prefer to use existing one instead of implementing one myself...
> > 
> > My best,
> >    J. Haataja
> > 
> > _______________________________________________
> > 3dem mailing list
> > 3dem at ncmir.ucsd.edu
> > https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem




More information about the 3dem mailing list