Borrar filtros
Borrar filtros

How to estimate the PSF from an blurred image?

15 visualizaciones (últimos 30 días)
Niranjan
Niranjan el 1 de Feb. de 2011
Comentada: Nisha Varghese el 20 de Dic. de 2020
Hi I am a newbie to image processing. I am working on Blind Image Deconvolution. I have a doubt regarding the PSF.How can one determine the unknown PSF in an blurred image and use the same to restore the image by blind deconvolution?? . Please clarify my doubt and correct me if I am wrong.Thanks in advance.
  2 comentarios
Daniel Armyr
Daniel Armyr el 22 de Feb. de 2012
Do you just have the blurred image, or do you also have a sharp version of the image? If you only have the blurred image, then you cannot find the PSF because you simply do not have enough information.
YE LING
YE LING el 4 de Ag. de 2019
Hi Daniel,
I do have a sharp image and a blur image, but how to get the PSF. I try to use the below equation to find it. But the result is not right. Can you help me with a code?
formula.png
Thanks in advance.

Iniciar sesión para comentar.

Respuestas (2)

Bjorn Gustavsson
Bjorn Gustavsson el 22 de Feb. de 2012
If you have some object that you "know should be a point", like for example a star, you could use the region around that as a start-guess for the PSF. If you dont have that you might have an object in your image with an edge that you "know should be a step" (preferably to the rows or columns) that will be blurred too (obviously) - but that edge-spread will be the Abel transform of the PSF. So you'd have to inverse-Abel it first.
But as Daniel pointed out you have a tricky problem.
HTH.
  2 comentarios
Rahim Gholami
Rahim Gholami el 4 de Jul. de 2018
Hi. I have the same problem.I have an image that is blurred but I couldn't remove the noise from it by guessing parameters of PSF. I also read some papers about calculating this parameters, but their theories were so hard to use in matlab. if anyone knows how i can estimate this parameters please help me. i don't have the original image and i have the blurred one only.
Matt J
Matt J el 4 de Jul. de 2018
Editada: Matt J el 4 de Jul. de 2018
There is no way to deduce the PSF without some information about the true object. How can one know if the "blur" seen in the image isn't in fact a true feature of the original object?

Iniciar sesión para comentar.


My@ma
My@ma el 3 de Jun. de 2019
Hello, please i need a help
I am curretly working on an image processing project, i have two images one blurred(SIMS image) and one sharp(TEM)
and i am trying to estimate PSF using the sharp one as an input and the blurred one as the output ,and after that using deconvolution to get the SIMS images sharpened
could you please give me method or propositions to let say automatically estimate PSF of this kind of system
thank you in advance
  2 comentarios
YE LING
YE LING el 4 de Ag. de 2019
Hi,
Do you find the way to find the PSF via the sharp and blur images? I also face this problem.
If you get the answer, please tell me.
Thanks in advance.
Nisha Varghese
Nisha Varghese el 20 de Dic. de 2020
Hi,
I also have the same problem. How to esimate PSF from blur and deblurred image pair (ie. I have original image and its blurred version, I have to find out PSF by using that pair.
Anyone could solve it?
Thanks

Iniciar sesión para comentar.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by