Settings for Particle Analysis

[FloKun]

Hello

With the help of microscope pictures of my o/w emulsion I am trying to get a PSD. My problem is that the particles nearly have the same colour as the background, the PSD is very broad and the particles are very close together (picture 1). To enhance the contrast I coloured my Oil with Sudan III (picture 2). It helps but I would prefer to relinquish Sudan III. Furthermore I dilute the emulsion to get fewer particles on each picture (3).

Nevertheless ImageJ does not do what I want. The particles are easy to see but I do not know how to tell that ImageJ. So if anybody knows some adjustments or tricks how to get my particles analysed, please let me know.

Thank you very much and greetings from Hamburg/Germany

[Michael Schmid]

Hi Florian,

here is a rough idea for segmenting it, at least the large ones. You  
will have to play with the settings (size range, circularity) of  
'Analyze particles' to eliminate the areas between some spheres.

selectWindow("2.jpg");  //assumes your original is called like this
run("8-bit");
run("Subtract Background...", "rolling=500 light");
run("Duplicate...", "title=edge-preserving_filter");
run("Thresholded Blur", "radius=2 threshold=30 softness=0.50  
strength=3"); //should be on the same line as run("Thresholded...
imageCalculator("Max create", "2.jpg","edge-preserving_filter");
setThreshold(0, 238);  //a bit critical, should select all spheres
run("Find Maxima...", "noise=20 output=[Segmented Particles] above");
//lower 'noise' might give you some more of the smaller spheres

Thresholded Blur is available at
   http://imagejdocu.tudor.lu/doku.php?id=plugin:start
It is used to eliminate the dark spots in the spheres.

The ideal way to do it would be a 3D Hough transform, but as far as I  
know the only Hough transform plugin is 2D (fixed size of circles,  
varying only the x and y coordinates). You would need to run it many  
times for all circle sizes.

Michael
________________________________________________________________

On 29 Jul 2010, at 17:19, Florian Kunde wrote:

> Hello
>
> With the help of microscope pictures of my o/w emulsion I am trying  
> to get a PSD. My problem is that the particles nearly have the same  
> colour as the background, the PSD is very broad and the particles  
> are very close together (picture 1). To enhance the contrast I  
> coloured my Oil with Sudan III (picture 2). It helps but I would  
> prefer to relinquish Sudan III. Furthermore I dilute the emulsion  
> to get fewer particles on each picture (3).
>
> Nevertheless ImageJ does not do what I want. The particles are easy  
> to see but I do not know how to tell that ImageJ. So if anybody  
> knows some adjustments or tricks how to get my particles analysed,  
> please let me know.
>
> Thank you very much and greetings from Hamburg/Germany
>
>
>
> <1.jpg><2.jpg><3.jpg>