Legacy ImageMagick Discussions Archive

As I mentioned above, typically Lucy and blind deconvolution do not work well with ideal motion blur or defocus. The only one that may work is Wiener. At least that is what I have I have used in my scripts.

http://ieeexplore.ieee.org/stamp/stamp. ... er=4408904

this is the paper, where they are doing blur removal of non uniform motion. but i don't know how it will work. code is not provided.

Sorry, I do not know how they do it and cannot access that paper.

Thanks for the reference. I will have to look at it in more depth later. But it requires multiple blurred input images and segments the images into regions, each of which have a separate motion blur, as best I can tell on quick look. The references in the paper are interesting and I will have to look at them also. But they have very specialized code and it is not very conducive to scripting in Imagemagick.

By the way, I believe Photoshop has some method of deblurring motion blur. But I have not tried to use it nor do they give any description of the algorithm that I have found. They seem to be able to estimate the motion blur by locating some point object and computing the psf from that minute image.

Here is a reference to work at Microsoft requiring two images.

http://research.microsoft.com/en-us/um/ ... RAPH07.pdf

You might also want to look at blind deconvolution such as at http://www.quarktet.com/joom/index.php/seddara

https://drive.google.com/file/d/0B6Eilz ... NBUk0/view

can i follow this method(---->above link)..will it work?

sorry try this link..
https://drive.google.com/file/d/0B6Eilz ... NBUk0/view
can i follow this method(---->above link)..will it work?

I cannot say. I have never used that method.

than you so much,
I have found the another method to restore the non uniform motion blur, but i am struggling to understand that.
Can you please explain me with your exposure?

I am not sure what you are asking? Can you point me to the reference to the non-uniform motion blur restoration? If something else, then please explain further what you need help to understand.

sorry i forgot to attach the link....
http://dx.doi.org.sci-hub.cc/10.1016/j. ... 013.05.001#
this is another method....they are telling total variation,shock filter,and salient structure. please give me the explanation on that. i will be ready for skype call also.

That reference looks like some good work. But I have not kept up with the literature to be able to understand fully what they are doing. Unfortunately, I do not have the time to dig into it and all the pertinent references to get up to speed.

Furthermore, even if I understood, I would not be able to script it with Imagemagick. Some one would have to write code.

Perhaps you can contact the authors and get some code from them.

i need some technical explanation of that work. can you please explain the work flow of that paper. then only i can start up the coding part...i will be ready to connect with you in personal-chat. please let me know when your are free.