How to pass from RBG image to CIE L*a*b* domain?

Hi everybody,

I want to convert an RGB image - composed from MRI sequences - to CIE L*a*b*. I used the rgb2lab built-in function but the output is quite different from expectations. I'm following the paper avaible from this link https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4709039/.

[Adam]

What are your expectations? Are the actual values in the LAB conversion wrong or are you expecting that plotting a LAB result in RGB to look different to what it does?

[Francesco vista]

I want to have the same L*a*b* image presents in the article in the link above.

[Adam]

They appear to use a pseudo-color mapping from greyscale to intensity that they can then presumably plot using a colourmap. I suspect you are missing this step, which you would need to read up about if you are not already familiar with pseudo-colouring. You can't just convert RGB to LAB and then plot the result with something like imagesc or imshow because it will simply interpret L as red, A as green, B as blue and give you an image that makes very little sense.

[Francesco vista]

Very thanks, I'm not familiar with pseudo-coloring so can you pass me some link in wich I can find some example or better a sequences of step to do in order to obtain the same result?

[Adam]

It's not something I've really done either, I just did a quick google search on it to understand on a basic level what it is.

[Francesco vista]

Great, i will read up about pseudo-coloring and I hope to fix it. Thank you for your time.