I have 3 different images and I have the data which is in frequency domain. The data size is : 192x256x88x32x3.

First 3 sizes are frequency domain dimensions
    4.dimension: channels
    5.dimension: different images' data in frequency domain

When I take FFT3D for this data and sum of squares in the channel direction I obtain the image. I want to normalize all this images such that their maximum image domain intensity value should be 1. The process I'm doing is taking 3D FFT for 32 channels. Then I'm finding maximum of the image intensity and multiplying the maximum value with frequency domain.

  1. Is my approach true?
  2. After taking absolute value, there are a lot of outliers. Will it be problem?
  • $\begingroup$ Can you please clarify what each dimension is exactly? When you say "first three dims are frequency domain dims", I understand 2 dims of spatial dimension plus a temporal dim. What does "32 channels" mean? Is that 32 sub bands of the spatial FFT? $\endgroup$ – A_A Oct 1 '16 at 9:19
  • $\begingroup$ This data is taken from MRI machine and each channel can monitor different sides of the brain. First three dimensions are spatial dimensions. $\endgroup$ – toygan kılıç Oct 2 '16 at 9:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.