Login  Register

Re: Integrated Density?

Posted by Ben Tupper on Aug 29, 2008; 12:52am
URL: http://imagej.273.s1.nabble.com/Integrated-Density-tp3695254p3695256.html

On Aug 28, 2008, at 12:01 PM, Kashif Zeeshan wrote:

> Hello experts,
>
> Can you kindly help me to understand the term "Integrated Density".  
> I read
> the documentation of ImageJ which says that this is the sum of all the
> pixels or equals to the product of Area and Mean Gray Value. But it  
> is very
> confusing for me. How we will interpret this? What is the  
> difference between
> density and integrated density in this regard? Is this term  
> "integrated
> density" is same as "density"? My aim is to compare the different  
> media
> giving different forms of fungi according to the density of the  
> pellets
> found in each medium and also if this is possible then I want to  
> classify
> the pellets in a medium according to the density i.e., I want to  
> classify
> pellets as less dense, average dense and dense.
> Any help will be highly appreciated.
>
> Kashif ZEESHAN

On Aug 28, 2008, at 7:22 PM, Robert Dougherty wrote:

> Kashif,
>
> Maybe this will help: density is an intensive quantity and  
> integrated density is an extensive quantity.  See  Wikipedia for an  
> explanation of extensive and intensive quantities.
>

Hi,

I find it a bit hard to untangle the questions posed.  But here are  
couple of thoughts...

(1) Be sure that you understand what the documentation actually says,  
"The sum of the values of the pixels in the image or selection."  In  
your case, it means that it sums the gray values of each pellet found  
- it is a measure of "grayness" on an individual particle or  
selection. (see http://rsb.info.nih.gov/ij/docs/menus/analyze.html#set )

(2) Your description of pellet counting could be used to measure a  
density of population - for example, X pellets per square cm. You can  
use ImageJ for this purpose.  For that you could analyze the image  
and retrieve either the number of pellets found.  Dividing the pellet  
count by the image area will give you the population density.  If you  
want to discriminate density patterns within an image, then you might  
want to dig deeper by retrieving the centroids of each pellet.  You  
could divide the image into quadrants of known area and get the  
population density in each quadrant, etc.

If I follow you correctly, I don't think you want the Integrated  
Density as you won't care how dark each pellet is.

Cheers,
Ben



Ben Tupper
Bigelow Laboratory for Ocean Science
180 McKown Point Road
POB 475
West Boothbay Harbor, ME 04575