3D Fourier Transform
Locked
 
|
Hi, guys!
I have a stack from a tomographic reconstruction that shows a big structure made of thin tubes/lines. These lines are present in all the structure, so I am wondering if it is possible to obtain a 3D Fourier Transform of the volume to see if there is some distance repetition. I have been making some tests, but what I have obtained is also a stack. It seems to be the 1D FT of each slice...and I was expecting a single image. Am I wrong? When you apply FT to a stack you also obtain a stack? If not, please help me to find the way! Thanks in advance! Andrea -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
|
|
Well Andrea,
what you actually get is a "2D FT of every slice" because ImageJ comes with 2D-FFT only, but there is the ImageJ PlugIn "FFTJ" that performs 3D FFTs. Why not have a look at the ImageJ PlugIn page? Here you are: <http://rsb.info.nih.gov/ij/plugins/fftj.html> and don' forget to read <http://rsb.info.nih.gov/ij/plugins/fftj2.html> Be aware that you use a power of two sized dimensions of your stack, otherwise the DFT applies and that will take longer to compute! HTH Herbie :::::::::::::::::::::::::::::::::::::::::::: Am 23.07.15 um 16:32 schrieb Andrea Chicano: > Hi, guys! > > I have a stack from a tomographic reconstruction that shows a big structure > made of thin tubes/lines. These lines are present in all the structure, so > I am wondering if it is possible to obtain a 3D Fourier Transform of the > volume to see if there is some distance repetition. > > I have been making some tests, but what I have obtained is also a stack. It > seems to be the 1D FT of each slice...and I was expecting a single image. > Am I wrong? When you apply FT to a stack you also obtain a stack? If not, > please help me to find the way! > > Thanks in advance! > > Andrea > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
|
|
In reply to this post by cromatina
Hi Andrea,
For a 3D version of FFT, have a look to P. Wendykier work : https://sites.google.com/site/piotrwendykier/software/parallelfftj Best, Thomas On 23/07/15 22:32, Andrea Chicano wrote: > Hi, guys! > > I have a stack from a tomographic reconstruction that shows a big structure > made of thin tubes/lines. These lines are present in all the structure, so > I am wondering if it is possible to obtain a 3D Fourier Transform of the > volume to see if there is some distance repetition. > > I have been making some tests, but what I have obtained is also a stack. It > seems to be the 1D FT of each slice...and I was expecting a single image. > Am I wrong? When you apply FT to a stack you also obtain a stack? If not, > please help me to find the way! > > Thanks in advance! > > Andrea > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html -- /***************************************************************/ Thomas Boudier, Associate Professor, UPMC, Université Pierre et Marie Curie, Paris, France. BioInformatics Institute (BII)/IPAL, Singapore. /**************************************************************/ -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
|
|
Hi Thomas,
thank you! This helps a lot! I'll check the work and try to apply to my samples. Best, Andrea 2015-07-24 6:52 GMT+02:00 Thomas Boudier <[hidden email]>: > Hi Andrea, > > For a 3D version of FFT, have a look to P. Wendykier work : > > https://sites.google.com/site/piotrwendykier/software/parallelfftj > > Best, > > Thomas > > > > On 23/07/15 22:32, Andrea Chicano wrote: > >> Hi, guys! >> >> I have a stack from a tomographic reconstruction that shows a big >> structure >> made of thin tubes/lines. These lines are present in all the structure, so >> I am wondering if it is possible to obtain a 3D Fourier Transform of the >> volume to see if there is some distance repetition. >> >> I have been making some tests, but what I have obtained is also a stack. >> It >> seems to be the 1D FT of each slice...and I was expecting a single image. >> Am I wrong? When you apply FT to a stack you also obtain a stack? If not, >> please help me to find the way! >> >> Thanks in advance! >> >> Andrea >> >> -- >> ImageJ mailing list: http://imagej.nih.gov/ij/list.html >> > > -- > /***************************************************************/ > Thomas Boudier, Associate Professor, UPMC, > Université Pierre et Marie Curie, Paris, France. > BioInformatics Institute (BII)/IPAL, Singapore. > /**************************************************************/ > > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
|
|
Hi Andrea,
you can also check out ImgLib2: http://fiji.sc/ImgLib2_Examples#Example_6c_-_Complex_numbers_and_Fourier_transforms <http://fiji.sc/ImgLib2_Examples#Example_6c_-_Complex_numbers_and_Fourier_transforms> It is part of Fiji and can be called through a Beanshell script for example. Hope this helps, Stephan > On Jul 24, 2015, at 10:48 , Andrea Chicano <[hidden email]> wrote: > > Hi Thomas, > > thank you! This helps a lot! I'll check the work and try to apply to my > samples. > > Best, > Andrea > > 2015-07-24 6:52 GMT+02:00 Thomas Boudier <[hidden email]>: > >> Hi Andrea, >> >> For a 3D version of FFT, have a look to P. Wendykier work : >> >> https://sites.google.com/site/piotrwendykier/software/parallelfftj >> >> Best, >> >> Thomas >> >> >> >> On 23/07/15 22:32, Andrea Chicano wrote: >> >>> Hi, guys! >>> >>> I have a stack from a tomographic reconstruction that shows a big >>> structure >>> made of thin tubes/lines. These lines are present in all the structure, so >>> I am wondering if it is possible to obtain a 3D Fourier Transform of the >>> volume to see if there is some distance repetition. >>> >>> I have been making some tests, but what I have obtained is also a stack. >>> It >>> seems to be the 1D FT of each slice...and I was expecting a single image. >>> Am I wrong? When you apply FT to a stack you also obtain a stack? If not, >>> please help me to find the way! >>> >>> Thanks in advance! >>> >>> Andrea >>> >>> -- >>> ImageJ mailing list: http://imagej.nih.gov/ij/list.html >>> >> >> -- >> /***************************************************************/ >> Thomas Boudier, Associate Professor, UPMC, >> Université Pierre et Marie Curie, Paris, France. >> BioInformatics Institute (BII)/IPAL, Singapore. >> /**************************************************************/ >> >> >> -- >> ImageJ mailing list: http://imagej.nih.gov/ij/list.html >> > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
|
|
Hi Stephan,
I'll check it! Thank you. Best, Andrea 2015-07-24 14:34 GMT+02:00 Stephan Preibisch <[hidden email]>: > Hi Andrea, > > you can also check out ImgLib2: > http://fiji.sc/ImgLib2_Examples#Example_6c_-_Complex_numbers_and_Fourier_transforms > < > http://fiji.sc/ImgLib2_Examples#Example_6c_-_Complex_numbers_and_Fourier_transforms > > > > It is part of Fiji and can be called through a Beanshell script for > example. > > Hope this helps, > Stephan > > > On Jul 24, 2015, at 10:48 , Andrea Chicano <[hidden email]> > wrote: > > > > Hi Thomas, > > > > thank you! This helps a lot! I'll check the work and try to apply to my > > samples. > > > > Best, > > Andrea > > > > 2015-07-24 6:52 GMT+02:00 Thomas Boudier <[hidden email]>: > > > >> Hi Andrea, > >> > >> For a 3D version of FFT, have a look to P. Wendykier work : > >> > >> https://sites.google.com/site/piotrwendykier/software/parallelfftj > >> > >> Best, > >> > >> Thomas > >> > >> > >> > >> On 23/07/15 22:32, Andrea Chicano wrote: > >> > >>> Hi, guys! > >>> > >>> I have a stack from a tomographic reconstruction that shows a big > >>> structure > >>> made of thin tubes/lines. These lines are present in all the > structure, so > >>> I am wondering if it is possible to obtain a 3D Fourier Transform of > the > >>> volume to see if there is some distance repetition. > >>> > >>> I have been making some tests, but what I have obtained is also a > stack. > >>> It > >>> seems to be the 1D FT of each slice...and I was expecting a single > image. > >>> Am I wrong? When you apply FT to a stack you also obtain a stack? If > not, > >>> please help me to find the way! > >>> > >>> Thanks in advance! > >>> > >>> Andrea > >>> > >>> -- > >>> ImageJ mailing list: http://imagej.nih.gov/ij/list.html > >>> > >> > >> -- > >> /***************************************************************/ > >> Thomas Boudier, Associate Professor, UPMC, > >> Université Pierre et Marie Curie, Paris, France. > >> BioInformatics Institute (BII)/IPAL, Singapore. > >> /**************************************************************/ > >> > >> > >> -- > >> ImageJ mailing list: http://imagej.nih.gov/ij/list.html > >> > > > > -- > > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
«
Return to ImageJ
|
1 view|%1 views
| Free forum by Nabble | Edit this page |