Hi,<br><br>Regarding the binary images, the way I get it is by loading a random dicom file, obtain the output (which is of class itk::Image). After that I just change a row (position and width set arbitrary) to a value of n (again arbitrary chosen) and the rest to n+1 in the data.
<br>Finally save the modified image into a new dicom file.<br><br>In another words I don't really have any phantom files.<br>Do mail me personally if you still want it, I'll try to rewrite the code to generate the file as I sort of overwritten my testing code...
<br><br>Regards,<br>Pechin<br><br><div><span class="gmail_quote">On 9/25/07, <b class="gmail_sendername">Jean-Pierre Roux</b> <<a href="mailto:jpr@creatis.insa-lyon.fr">jpr@creatis.insa-lyon.fr</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Pechin Lo wrote:<br>> I'm using ITK 3.2.0<br>> Should I switch to 3.4 and try again?<br><br>I'm afraid it will not change a lot.<br>In the very last version a gdcm, for a few images, we have, after<br>compression/decompression a few pixels that differs from +/-1 ...
<br>Could you send me one of your 'binary images' that only consists of n<br>and n+, to see if we can track the bug?<br>Thx<br>Jean-Pierre Roux<br>><br>> On 9/25/07, *Mathieu Malaterre* <<a href="mailto:mathieu.malaterre@gmail.com">
mathieu.malaterre@gmail.com</a><br>> <mailto:<a href="mailto:mathieu.malaterre@gmail.com">mathieu.malaterre@gmail.com</a>>> wrote:<br>><br>> On 9/25/07, Steve M. Robbins <<a href="mailto:steve@sumost.ca">
steve@sumost.ca</a><br>> <mailto:<a href="mailto:steve@sumost.ca">steve@sumost.ca</a>>> wrote:<br>> > On Mon, Sep 24, 2007 at 11:50:05PM +0200, Pechin Lo wrote:<br>> > > After some tweaking around, I found out that JPEG compression
<br>> works provided<br>> > > that the minimum difference between two pixels is 2.<br>> > > For example, if I have an image that only consists of n and<br>> n+1, the saved<br>> > > image will only contains the value n.
<br>> > > However, if the image consists of value n and n+2, then the<br>> saved image will<br>> > > contains both n and n+2.<br>> > > In other words, the image will stores fine if I multiply it by
<br>> a factor of<br>> > > 2.<br>> > ><br>> > > Can someone please tell me why is this happening?<br>> ><br>> > I'm not familiar with the code in question, but is it possible
<br>> > that you're using a lossy JPEG compression?<br>><br>> Hi all,<br>><br>><br>> Could you please let me know if you are using ITK 3.2 / ITK 3.4<br>> or CVS ?<br>><br>> Thanks
<br>><br>> --<br>> Mathieu<br>> Ps: 3.4 and CVS should *not* produce this weird behavior.<br>><br>><br>><br>><br>> --<br>> Pechin Lo Chien Pau<br>> Ph.d. Student<br>> Department of Computer Science (DIKU)
<br>> University of Copenhagen<br>> Universitetsparken 1, Office S226<br>> DK-2100 Copenhagen, Denmark<br>> Phone: +45 3532 1433<br>> ------------------------------------------------------------------------
<br>><br>> _______________________________________________<br>> Insight-users mailing list<br>> <a href="mailto:Insight-users@itk.org">Insight-users@itk.org</a><br>> <a href="http://www.itk.org/mailman/listinfo/insight-users">
http://www.itk.org/mailman/listinfo/insight-users</a><br>><br></blockquote></div><br><br clear="all"><br>-- <br>Pechin Lo Chien Pau<br>Ph.d. Student<br>Department of Computer Science (DIKU)<br>University of Copenhagen<br>
Universitetsparken 1, Office S226<br>DK-2100 Copenhagen, Denmark<br>Phone: +45 3532 1433