While performing shading correction in ImageJ to remove local fluorescent light intensity variations due to camera issues, most sites recommend dividing the uncorrected image by the base image and then multiplying by 255. I am just confused as to why we are dividing here? Why not subtract the uncorrected image by the base image?
I think that sites tried to speak you to do "split image" in ImageJ
When subtracting you also subtract data (pixel intensity) from your region of interest (ROI) i.e. from your image. When you divide (e.g. by the average background number), if your pixel data in your ROI is higher than that number for the average background, you will not get zero (this is important for making a MASK).
To make a MASK:
1. Divide your Image by the average background intensity (keep an open copy of the original for later - let's call this Orig_Im).
2. Multiply resultant DIVIDE image by 255 (to cause all pixel intensities to saturate or become 255 - be sure to be using 8 bit images).
3. Perform a Logical AND operation. By using the Image J/Figi "Process/Image Calculator/AND" tool you can then perform a logical AND operation of your Orig_Im with your MASK. The resultant image should appear to be background subtracted, but the crucial difference is that the pixel intensities in your original image ROI should be UNCHANGED (unlike following a simple subtraction).
Although there are fancier ways of doing this, the trick is to make your MASK by trial and error (i.e. by dividing by different numbers). If the MASK is made appropriately, this will remove the background but will not alter pixel values in your image - which you could/should verify. Compare this to simply subtracting the background - you will see that the pixel intensities of your ROI are reduced by a background subtraction but not by the AND operation using your MASK. Hope this helps.
Hope this helps.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- RNA