Define edges

[retgar]

Hello, i`m requesting some help please!
Cant find the way to define the edges, so that imageJ "analyze particles"
can analyze each separate particle.
"find edges' and threshold doesnt seem to help. also tried 'FeatureJ' and
'image edge' plugins
https://imgur.com/a/dE26Zcj



--
Sent from: http://imagej.1557.x6.nabble.com/

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

[Joel Sheffield]

I took the image you posted, and ran the following:

run("8-bit");
setAutoThreshold("Default");
//run("Threshold...");
run("Create Mask");
run("Fill Holes");
run("Watershed");
selectWindow("Clipboard");
run("Restore Selection");
run("Analyze Particles...", "size=500-Infinity show=Outlines display
include summarize");

The problem with this sample is that the central portion of each object is
bright, and so you can't threshold the entire particle. Each particle is
thresholded with a hole.   Instead, you can convert to a mask, and then use
the Binary options to (a) fill holes, and (b) watershed, to isolate
particles that touch.  Then you create a selection around the masked
particles, and restore it to your original image.  then you can run Analyze
particles, and check "includeholes" in the options.  You also have to set a
lower bound for the particles so that you don't pick up the small
detritus.  It's not perfect --note the cluster at 8 o'clock, but it's
pretty good.  How much precision do you need?

[image: image.png]

Joel B. Sheffield, Ph.D
Department of Biology
Temple University
Philadelphia, PA 19122
Voice: 215 204 8839
e-mail: [hidden email]
Office:  Biolife 311
URL:  *https://bio.cst.temple.edu/~jbs/ <https://bio.cst.temple.edu/~jbs/>
<http://tinyurl.com/khbouft>*


On Thu, Apr 4, 2019 at 1:24 PM retgar <[hidden email]> wrote:

> Hello, i`m requesting some help please!
> Cant find the way to define the edges, so that imageJ "analyze particles"
> can analyze each separate particle.
> "find edges' and threshold doesnt seem to help. also tried 'FeatureJ' and
> 'image edge' plugins
> https://imgur.com/a/dE26Zcj
>
>
>
> --
> Sent from: http://imagej.1557.x6.nabble.com/
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

[retgar]

Hello, and first of all thank you for such promising result!
I have tried your method and results was really good, but still having a
little downsides sometimes.
Like for example, watershed sometimes crop one patricle on two halfs or
doesnt separate some particles at all
<http://imagej.1557.x6.nabble.com/file/t382279/11.jpg>
Also, i tried in a such way:
8-bit
subtract background
window/level - auto (or without this option whatsoever)
threshold
Fill holes
watershed

But one thing that i noticed, is that thresholding values sometimes may
vary, and i guess there is no way to do it automaticaly? And to answer your
question: i`m studying grow kinetics, so precision is the key, in order to
obtain sustainable results



--
Sent from: http://imagej.1557.x6.nabble.com/

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