Inclusion of blurred/hazy images in Analysis?


Hi team,
I’m quite through with the CP by now, especially with yr help.

My image acquisition includes 25 images (5x5) per well of 96-well plate.
Now, few of them are bit hazy/blurred in nature.

If I include these blurred images alongwith the other gud ones into the Illumination correction pipeline, How much its going to effect the overall Illumination function.
If these needs to be excluded, is there anyway of automation which can check the blurredness so that I can treat it from well-to-well in a high-throughput manner.

Kindly let me know.



The second question I can answer right away: This is on our list of features to include. There is definitely a need for image quality control for high-throughput screens, and the ability to do so prior to the screen itself rather than retrospectively would be a great asset. We’re working on putting a pipeline together which does this task, but it won’t be done for a while yet.

In the meantime, I suggest adding the MeasureSaturationBlur module to your pipeline. Have a look at this previous post: viewtopic.php?f=2&t=120&p=557&hilit=MeasureImageSaturationBlur#p557. It’s a bit dated, but should be helpful.

I will check on the answer to the first question and get back to you.



An answer to your second question: if the proportion of your blurred images is relatively small compared to your total number of images, your illumination correction should be fine. Even if not, unless the intensity of your blurred images is very nonlinear, it shouldn’t be a problem since the overall illumination function should be fairly smooth anyway (or you should adjust the module parameters to make it so).



Thanks Mark.

I guess, the number of blurred images are small.
Let me work with, what you have suggested.
Mridul KK