http://imagej.273.s1.nabble.com/Re-Chromatin-condensation-tp5013860p5013869.html
This is the contrast of the co-occurrence matrix, not the contrast of the image. Increasing the step size might reverse the trend.
> -----Original Message-----
> From: ImageJ Interest Group [mailto:
[hidden email]] On Behalf Of
> Lendl,Thomas
> Sent: Monday, August 3, 2015 15:44
> To:
[hidden email]
> Subject: Re: Chromatin condensation
>
> Hi Oli,
>
> Thanks a lot for your help, this looks promising. I'm just not sure how to
> interpret the result because It's somehow counterintuitive to have a lower
> contrast in more structured objects. Anyway I'll do a testrun on a couple
> more images to see if the output is consistent.
>
> Best,
> Thomas
>
> -----Original Message-----
> From: ImageJ Interest Group [mailto:
[hidden email]] On Behalf Of
> Burri Olivier
> Sent: Montag, 03. August 2015 14:50
> To:
[hidden email]
> Subject: Re: Chromatin condensation
>
> Hi Thomas,
>
> I had some success in calibrating the chromatin condensation using the
> Texture Analyzer
http://rsb.info.nih.gov/ij/plugins/texture.html>
> On your images:
> Cropped the Z stacks to contain slices with nuclei Ran the GLCM Texture
> plugin with a step size of 3 pixels
>
> Looked at the Average Contrast output of each image and got these values
>
> Control 229.209
> Mid 205.782
> Condensed 135.794
>
> What I would suggest is try different step sizes to find the one that
> maximizes the delta between the conditions and perhaps use that.
>
> You can also try the other metrics from the plugin, which have a similar trend
> (Except for Correlation, which makes sense);
>
> Best
>
> Oli
>
> --
> ImageJ mailing list:
http://imagej.nih.gov/ij/list.html>
> --
> ImageJ mailing list:
http://imagej.nih.gov/ij/list.html