Posted by
Jeremy Adler on
Oct 07, 2014; 5:16am
URL: http://imagej.273.s1.nabble.com/Timelapse-Noise-Quantification-tp5009923p5009926.html
Noise is a tricky subject - clearly if you know what the signal should be then you can measure the noise, but usually we don't.
An alternative that we use in colocalization measurements is based on the idea that noise can be addressed by acquiring two images of the specimen- in the absence of noise and movement they should be identical. The magnitude of the noise can be displayed using a scattergram and quantified using the correlation between the two nominally identical images.
Adler J., Bergholm F., Pagakis S.N., Parmryd I (2008)
Noise and Colocalization in Fluorescence Microscopy: Solving a Problem.
Microscopy & Analysis, Sept. 2008, 7-10.
J. Adler, S.N. Pagakis and I. Parmryd (2008)
Replicate Based Noise Corrected Correlation for Accurate Measurements of Colocalization
J. Microscopy 230(1),121-133.
________________________________________
From: ImageJ Interest Group [
[hidden email]] on behalf of Rebecca Keller [
[hidden email]]
Sent: 06 October 2014 20:11
To:
[hidden email]
Subject: Timelapse Noise Quantification
Dear Imagers,
I have done some experiments adding Gaussian noise to my time lapses to
determine the efficacy of a certain filter, and would like to give an idea
of how much noise had been there originally. Any suggestions, then, for
measuring this "baseline experimental noise?"
I am thinking perhaps I should take a z-projected (well, "t-projected,"
really) mean image of the series, subtract it from each image in the stack,
then do a z-projected standard deviation, report the average value of this
projected image? Or instead of subtracting the first mean image, maybe I
should divide the stack by it?
Any advice here would be welcome,
Jacob Keller
--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html