Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > sci.image.processing > #4168

Re: MOtion deblur

From Martin Leese <please@see.Web.for.e-mail.INVALID>
Newsgroups sci.image.processing
Subject Re: MOtion deblur
Date 2016-10-31 09:20 -0600
Organization A noiseless patient Spider
Message-ID <nv7nfh$k9m$1@dont-email.me> (permalink)
References <bfb8143c-fc6c-489b-9933-7ea0f3305413@googlegroups.com> <nutc3m$l7q$1@dont-email.me> <c92ed908-121e-46d8-8751-dcdd00a10343@googlegroups.com>

Show all headers | View raw


pdaraja@gmail.com wrote:
> On Thursday, October 27, 2016 at 10:35:22 PM UTC+5:30, Martin Leese wrote:
>> pdaraja wrote:
...
>>  > in addition to that why the deconvolution
>>  > algorithm requires the simulating of
>>  > motion blur to restore the deblurred image.
>>
>> It does not.  However, you do need to
>> estimate the blurring function, Blur(s).
>> For constant velocity blur typically this
>> would be its direction and length.
> 
> thank you so much martin.
> i have one big doubt, if i have a blurred 
> image as a input, how can i determine blur 
> kernel and how can i restore it as 
> deblurred.

How you determine Blur(s) depends on the
type of blur.  If the motion is at constant
velocity then you just need its direction
and length.  One way to do this is to
examine what would have been point sources
in the unblurred image.  These are blurred
to lines (with a length and direction).

You can then restore the image by applying
the deconvolution algorithm, making sure to
do something about the infinities.

-- 
Regards,
Martin Leese
E-mail: please@see.Web.for.e-mail.INVALID
Web: http://members.tripod.com/martin_leese/

Back to sci.image.processing | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread


Thread

MOtion deblur pdaraja@gmail.com - 2016-10-26 23:57 -0700
  Re: MOtion deblur Martin Leese <please@see.Web.for.e-mail.INVALID> - 2016-10-27 11:05 -0600
    Re: MOtion deblur pdaraja@gmail.com - 2016-10-30 22:40 -0700
      Re: MOtion deblur Martin Leese <please@see.Web.for.e-mail.INVALID> - 2016-10-31 09:20 -0600
        Re: MOtion deblur pdaraja@gmail.com - 2016-11-01 00:25 -0700
          Re: MOtion deblur pdaraja@gmail.com - 2016-11-01 00:26 -0700
          Re: MOtion deblur Martin Leese <please@see.Web.for.e-mail.INVALID> - 2016-11-01 09:37 -0600
            Re: MOtion deblur pdaraja@gmail.com - 2016-11-02 23:12 -0700
              Re: MOtion deblur Martin Leese <please@see.Web.for.e-mail.INVALID> - 2016-11-03 08:51 -0600
                Re: MOtion deblur pdaraja@gmail.com - 2016-11-03 21:43 -0700
                Re: MOtion deblur dale <dale@dalekelly.org> - 2016-11-04 21:08 -0400

csiph-web