https://www.researchgate.net/publication/235662494_On_separating_overlapped_fingerprints [Paper name-"On separating overlapped fingerprint", author-Fanglin Chen,Jianjiang Feng,Jie Zhou] I am unable to understand the process of local fourier analysis in the initial orientation field estimation. I have tried searching over the internet but in vain. enter image description here

The sentence I am unable to understand has been marked in the paper. What is this 16 x 16 and then again from where do i get this 64 x 64? I can apply a gaussian filter by taking each pixel as centre. But how can i do that taking a 16 x 16 block as centre? The meaning of that sentence is absolutely unclear to me. I have looked at the original paper from where this has been taken and also this paper (https://www.researchgate.net/publication/261469993_Automatic_segmentation_of_latent_fingerprints) [Paper-Automatic Segmentation of Latent Fingerprint author-Anil K.Jain] but things are very unclear to me. Please, I would request the community to kindly help me. I am stuck at this for the last 2 weeks. Thanks in advance to anyone who will help me

  • $\begingroup$ Please provide a full reference, including title and authors of the paper, so that others with a similar question can find this via search. We have collected some advice here. Thank you! $\endgroup$
    – D.W.
    Sep 30 '20 at 20:41
  • $\begingroup$ Please transcribe text and mathematics, so the question can be found via search and so that this is accessible to individuals with visual impairments. Note that your "this paper" is not going to work for anyone else. Please proof-read your post. $\endgroup$
    – D.W.
    Sep 30 '20 at 20:42
  • $\begingroup$ (Can you please include /D:?) $\endgroup$
    – greybeard
    Oct 3 '20 at 20:47
  • $\begingroup$ (To me, 16×16 pixels looks arbitrary. I take an 64×64 [pixel] window to be centred "at each block" - which includes 9 blocks completely, one half of 12 blocks and one quarter of 4 more.) The title of the first paper lacks the trailing/plural s. The importance of such detail can be seen in the second reference: I don't have access to your "device D:", and title and author as stated give diverse search engine near misses. $\endgroup$
    – greybeard
    Oct 3 '20 at 21:15
  • $\begingroup$ Many many thanks for replying @greybeard sir. But , I would like to know, how you can centre on a block and multiply the local image in 64 x 64 window with gaussian function...I am unable to visualize it. Gaussian filtering would require centering on a pixel for a block of certain size and that block size needs to be odd, I do not know how to filter by centering on a block. If you can kindly provide further explanation or point out relevant resources, it would be of great help. $\endgroup$ Oct 5 '20 at 3:17

This site is temporarily in read only mode and not accepting new answers.

Browse other questions tagged .