[vmtk-users] convert .DAT file to a .VTP file

[Haim Ezer]

Hello, Luca, everybody.

I wonder if there is a simple way, to convert many files, saved by VMTK in dat format, to the vtp format.

I saved about a 100 centerlines, in the dat format, for the use in matlab, and now I want to use it in VMTK.

VMTK doesn't seem to read this file format, although it saves in this format.

Thanks

Haim

 

--

Haim Ezer, MD
Department of Neurosurgery,

The western Galilee hospital in Nahariya

Israel

E mail: haim...@gmail.com ,haim...@yahoo.com

[Luca Antiga]

Hello Haim,

 as you know the dat format throws away the cell connectivity information (it only retains the point coordinates), 

so there's not generic way of recreating a vtp from it.

However, as in your case, we can probably infer the connectivity by the fact that lines are written as a sequential

list of points.

One question before I can suggest you a solution: how many lines are there per file (i.e. do your dat files contain

once centerline per file or do you have more?)

Best,

Luca

------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev_______________________________________________
vmtk-users mailing list
vmtk-...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/vmtk-users

--

Luca Antiga, PhD

Cofounder and Principal Scientist, Orobix Srl

via L.A. Muratori 3, 24123 Bergamo, Italy

orobix: www.orobix.com

home: lantiga.github.com

twitter: twitter.com/lantiga

[Ieva Vaišnoraitė-Navikienė]

Hi Luca,

I also had a similar problem (file format without cell connectivity information). There exists many points for many centerlines in one file. Maybe you have an idea how I can convert that file to vtp?

Thank you for your time,
Ieva

[Luca Antiga]

Hi Leva,

I don’t know if this is true in your case, but usually it is reasonable to assume that between the end of one and the beginning of another the distance between points is greater than the average distance between two points belonging to the same centerlines.

You can write a small Python code that does the parsing line by line and builds a vtkPolyData centerline, creating a new line.

Reading a file: 

with open('data.txt', 'r') as f:
data = f.readlines()

# data is now a list of text lines.

import vtk

points = vtk.vtkPoints()

points.SetNumberOfPoints(len(data))

cells = vtk.vtkCellArray()

radii = vtk.vtkFloatArray()

radii.SetName(‘MaximumInscribedSphereRadius’)

radii.SetNumberOfTuples(len(data))

centerlines = vtk.vtkPolyData()

centerlines.SetPoints(points)

centerlines.SetLines(cells)

centerlines.GetPointData().AddArray(radii)

distance_threshold = 0.1

cellpts = []

prev = None

for i, line in enumerate(data):

x, y, z, r = [float(el) for el in line.split()[:4]]

if not prev:

prev = [x,y,z]

id = points.SetPoint(i,x,y,z)

radii.SetValue(i,r)

        if vtk.vtkDistance2BetweenPoints([x,y,z],prev)**0.5 > distance_threshold:

cells.InsertNextCell(len(cellpts))

for ptid in cellpts:

cells.InsertCellPoint(ptid)

cellpts = []

cellpts.append(id)

prev = [x,y,z]

writer = vtk.vtkXMLPolyDataWriter()

writer.SetInput(centerlines)

writer.SetFileName(‘foo.vtp’)

writer.Write()

Disclaimer: I just banged out the code in my email, I don’t guarantee it will just work, but you get the idea.

Luca

--
You received this message because you are subscribed to the Google Groups "vmtk-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vmtk-users+...@googlegroups.com.
To post to this group, send email to vmtk-...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.