Re: quantify the fibers with the red color

Posted by pang on
URL: http://imagej.273.s1.nabble.com/quantify-the-fibers-with-the-red-color-tp5003734p5003741.html

[cid:image001.png@01CE773C.911CCD10]



I used the following macro to quantify the area. There is some artifact due to the edge circle,



run("Split Channels");



selectWindow("image4072-001.jpg (green)");  // I like to use green as it is darkest in red, providing best contrast.

setAutoThreshold("Default dark");

//run("Threshold...");

run("Measure");



run("Subtract Background...", "rolling=50 light");



setThreshold(155, 227);

run("Measure");



My estimation is 121097/1044914, about 11% of area (partial circle).

121097 is the pixel number for the red color, 10440914 is the pixels within the circle.





-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of John Hayes
Sent: Tuesday, July 02, 2013 3:41 PM
To: [hidden email]
Subject: Re: quantify the fibers with the red color



Hi,



Can you tell us what you need to quantify about the dense red fibers in the image? The number of reddish pixels, the length of the fibers in the imaging field, etc.???



In any case, I recommend starting by splitting the RGB into separate channels, and thresholding the red channel to your liking to identify what you think is signal and exclude the noise... Then, « quantify » based on whatever is physiologically-relevant to your case, but it's unclear what that is right now.



Best,



John



Le 2 juil. 2013 à 21:32, Thuraya Medhat a écrit :



> i need to quantify the dense red color in the image please help me...

>

> --

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

> <image4072-001.jpg>



--

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

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