Posted by
Michael Schmid on
Nov 15, 2007; 1:46pm
URL: http://imagej.273.s1.nabble.com/Area-filters-when-counting-objects-on-an-image-tp3698002p3698003.html
Hi,
one thing you could try is creating a mask (Edit>Selection>Create Mask),
then runProcess>Binary>Watershed.
You could also clean the shapes e.g remove the legs or antennae,
by Process>Binary>Open before Watershed or the like.
Analyze particles can limit the area of the particles counted
or count only particles with a given circularity range (see the
documentation).
An alternative might be writing various values to the results
table (Analyze>Set Measurements) and filtering by an ImageJ
macro or a separate program.
Michael
________________________________________________________________
On 15 Nov 2007, at 12:59, Iñaki Etxebeste Larrañaga wrote:
> Dear ImageJ users,
>
> I’m an ex – ImageTool user, who is intending to use ImageJ to count
> beetles
> on pictures taken to samples.
>
> I downloaded ImageJ not so long ago, three days, and so far I’ve
> managed to
> count samples where the objects are completely separated. I adjust the
> threshold and then analyse the particles. No problem with that, but
> when the
> objects touch each other, or their out of a certain size range (ok,
> I can
> say I want a certain range of areas to be counted, but I may have
> beetle
> "rests" in my sample, that I need to analyse.
> In ImageTool one could apply a filter (save the filter), and
> classify the
> groups of objects regarding their size, so I could get a quite
> accurate
> estimate of the number of beetles in that sample, even if they were
> overlapped. This program even gave the display of the filtered
> sizes and so
> on. But it was quite difficult to export data, and gave errors on
> my PC, so
> I've decided to switch, but now I miss this options... Is there a
> way to
> apply such filters? Any other suggestions to carry such analyses?
>
> I guess this kind of questions are posted quite often, when I've
> searched
> "count" on the mail list, I get quite a load of results... sorry if
> I had
> not found the answer yet, but I'm in a kind of hurry...
>
> Well, thanks a million in advance,
> Greetings
>
>
> Iñaki Etxebeste Larrañaga
> M.Sc. Biologist
> E T S II AA Universidad de Valladolid
> Avda. Madrid 57
> 34071 Palencia (Spain)
>
>