Extrapolating Quantitative Data from Topographic Picture

I have limited ImageJ experience so I do not know where to start looking for helping with this. I need to extract (if possible) concentration of a particular element from the picture attached. The scale bar on the side can be converted from counts (current units) to % (concentration units). Any ideas/thoughts? Please let me know if you need any more information. Thanks in advance.

Reconstructing quantitative data from false-colored RGB images can be tricky, so it’s always a better option to contact the author of the image and ask for the original data.

Nevertheless, if you don’t have access to the original data, you can try the following to get an approximation of the original values using the calibration bar in your image:

  • Create a line selection on your calibration bar from 0 to 80 (while holding the Shift key to keep it vertical)
  • Convert the image to HSB color space (Image > Type > HSB Stack)
  • Use Analyze > Set Scale… to set the known length of the line selection to 80 (this will make the following profile plot scale correctly)
  • For each slice (i.e. Hue, Saturation, and Brightness), plot an intensity profile by pressing K (or Analyze > Plot Profile)

This will give something like:




From these plots, you should be able to derive a formula (or several formulas) for the different segments, allowing you to reconstruct the original signal value from the H,S and B values of each pixel. You can use the Image Expression Parser for that.

In case you are creating false-colored images yourself in the future, I recommend to use a color table (LUT) like mpl-viridis (Image > Lookup Tables > mpl-viridis), as this one allows to reconstruct the signal from the brightness channel only, without the need to take into account hue and saturation:

As illustrated below, mpl-viridis remains interpretable even after converting an RGB image to grayscale:

And here’s the same using the Spectrum LUT:

1 Like