Wavelength to RGB coversion

Posted by Daniel James White on Aug 23, 2010; 12:01pm
URL: http://imagej.273.s1.nabble.com/Wavelength-to-RGB-coversion-tp3687192.html

Hi Esteban,

On Aug 7, 2010, at 6:00 AM, IMAGEJ automatic digest system wrote:

> Date:    Fri, 6 Aug 2010 12:21:59 -0700
> From:    "G. Esteban Fernandez" <[hidden email]>
> Subject: Wavelength to RGB coversion
>
> I have 4D (xyz + lambda/spectral) stacks from a Zeiss 710 confocal that I
> need to render in 3D, with the lambda dimension pseudocolored in RGB.
> Zeiss AIM/LSM and ZEN software do not render 3D volumes of lambda-colored
> z-stacks or even export series of lambda-colored planes to TIFF (I'd have to
> do 'em one by one)

there is a very good reason why it does not do that.....


> so I'm going to color each channel according to
> wavelength using ImageJ's Image5D format.  Might there be an ImageJ plugin
> or other software that reads the metadata and does this automatically?
>
> Thanks,
> Esteban

The problem here is that there is no reliable/robust way to convert a wavelength into an RGB colour.

1) what will you do for far  red wavelengths that are invisible to the eye?
        Have them be invisible?
        Use some other arbitrary  colour... which you then cant use for the wavelength that looks most like it?

2) Similarly in the UV, you have a problem also. It has no human perceived colour to map to.

3) Worse still our eyes are most sensitive to Green, less to to red, and worst to Blue.
        This is biology - not much you can do about it.
        It mean that you can not compare the brightness of a green pixel with that of a blue one,
                even when they have the same intensity value, the green will "look" brighter.

        This basically makes a nonsense of the whole idea of trying to "see" all the wavelengths at the same time.
        There is simply too much in a spectral, xyz,  scan for our eyes and brain to deal with in any useful way
        (other than just looking pretty - art not science)

A good read of
http://en.wikipedia.org/wiki/Color_vision
and other similar pages might help you get the idea.


What you really need to do is some data reduction trick,
where you throw away info that is not interesting, abut keep the stuff that you want to see.
Its about seeing the wood, despite the trees.

As is often the case,
your actual problem here is not on trying to do the multi wavelength visualisation
(which, as I explain above, will not give you a final image that you are able to easily interpret scientifically)
but rather that you need to define better what it is that you want to actually visualise,
and reduce the dimensionality of the data and its complexity to such a state
that you can visualize what it is that you are after.

if you want to chat about it off list, feel free to contact me!

cheers

dan





Dr. Daniel James White BSc. (Hons.) PhD
Senior Microscopist / Image Visualisation, Processing and Analysis
Light Microscopy and Image Processing Facilities
Max Planck Institute of Molecular Cell Biology and Genetics
Pfotenhauerstrasse 108
01307 DRESDEN
Germany

[phone] (German Mobile)
[phone] (Work phone at MPI-CBG)
[phone] (Fax MPI-CBG LMF)

http://www.bioimagexd.net  BioImageXD
http://pacific.mpi-cbg.de                Fiji -  is just ImageJ (Batteries Included)
http://www.chalkie.org.uk                Dan's Homepages
https://ifn.mpi-cbg.de  Dresden Imaging Facility Network
dan (at) chalkie.org.uk
( white (at) mpi-cbg.de )