Processing a video

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Processing a video

Mahdi Razaz
Hi Everyone

I'm quite new to to image processing world and trying to find a way to count a lot of bubbles in a series of videos (see yhe link below).

The purpose of this email is to seek some help and see see if it is possible to isolate the bubbles in the sample frame or no. With my experience and level of expertise, it is not.

Thanks for your help,

https://www.dropbox.com/s/6s3426by30xivkl/Movie%201.mov?dl=0

--
-------------------------------------------------
Mahdi Razaz ([hidden email]<mailto:[hidden email]>)
Postdoc Fellow
Dept. of Marine Sciences, University of Georgia
247 Marine Sciences Building, Athens GA 30602
voice: 706 542-2405
Website: http://www.marsci.uga.edu/directory/mahdi-razaz

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

1-0904090252B.JPEG (445K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Processing a video

Michael Schmid
Hi Mahdi,

Assuming that the bubbles you want to count are the dark objects quickly
moving up near 1/3 from the left:

In theory it should work like the following:

First you need to import the movie as an ImageStack. If ImageJ can't
open the movie, use a tool like ffmpeg to convert it into single images
and then get those with Import>Image Sequence.
If the amount of data is too large, crop it to the relevant area.

It seems that the background does not change much, so an operation like
Image>Stacks>Z project should give you the background. Best use
'Median'. If you have black bubbles, 'Max' would be better, but the
white objects floating around would disturb it.

Then use the Image Calculator to calculate the z-projected image minus
the original stack (subtract; since the original is darker, it has to be
the second argument). Finally, use a plugin like TrackMate to track the
bubbles.


The first problem with your image is that the dark bubbles are invisible
in front of the dark objects in the background, so tracking the bubbles
will be very difficult or impossible. You would need a uniform white
object behind the track of the rising bubbles (e.g. white sheet; avoid
shadows of the bubbles on it).

Another problem is that the dark bubbles appear not well-defined, some
of them are very blurry and it is not clear whether these objects are
single bubbles or whatever. As a rule of thumb, if it is difficult to
decide for a human what is a valid object and what isn't (or how many
valid objects a blob contains), there is almost no chance to get it done
by image processing.

So, with the current movie it looks rather hopeless to me.

Michael
________________________________________________________________
On 25.10.18 17:41, Mahdi Razaz wrote:
 > Hi Everyone
 >
 > I'm quite new to to image processing world and trying to find a way
to count a lot of bubbles in a series of videos (see yhe link below).
 >
 > The purpose of this email is to seek some help and see see if it is
possible to isolate the bubbles in the sample frame or no. With my
experience and level of expertise, it is not.
 >
 > Thanks for your help,
 >
 > https://www.dropbox.com/s/6s3426by30xivkl/Movie%201.mov?dl=0

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: Processing a video

Alan Hewat-2
Michael's solution for background subtraction looks good, but perhaps Mahdi
should explain the physical situation. Does he only have these videos, or
can he intervene eg to place a white card behind the upper part of the
stream ? But he will still have varying bubble size, clumping, bubbles
behind bubbles etc.

If these are gas bubbles, measuring their average size and simply
collecting the gas would give a good estimation of their number :-)

Alan


--
______________________________________________
Dr Alan Hewat, NeutronOptics, Grenoble, FRANCE
<[hidden email]> +33.476.98.41.68
        http://www.NeutronOptics.com/hewat
______________________________________________

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html