Re: Assessing bird colouration

[Ross Ahmed]

Many thanks for all the comments so far.

 

I'll just quickly summarise the background to my query. In late November, a
Hoopoe was found dead in Durham, UK and is now stored at the Discovery
Museum in Newcastle, UK. To my eyes, the dark colouration of the orange on
the bird suggests it may belong to the subspecies saturata (which originates
from Asia), as opposed to the subspecies epops (which originates from
Europe). As concluded by Ericson, a key difference between these subspecies
is the colouration of the orange. A photo of the bird can be found here:
http://www.flickr.com/photos/31121749@N06/5276824748/

 

If it is possible to produce equivalent results as Ericson, this would help
in deciding which subspecies the bird belongs to (and therefore whether it
may have originated from Asia or Europe).

 

So could anybody provide a step-by-step method, which would produce results
comparable with those of Ericson, for photographing the bird (my camera is a
Canon 40D) and analysing the photos in Image J (or any other program)?
Robert Dougherty in particular hinted that this may be possible.

 

Thanks again

 

Ross

 

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
Wolfgang Schechinger
Sent: 20 December 2010 06:51
To: [hidden email]
Subject: Re: Assessing bird colouration

 

If you can *see* the differences between the birds with your eyes in
pictures taken with a camera that takes RGB images, I think, it should be
possible. Maybe you need some sort of "standard curve" / reference data sets
created by visual assertion of birds to groups.

Actually this might be a good job for a neuronal network.

 

Have fun

 

Wo

--

GRATIS! Movie-FLAT mit über 300 Videos.

Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome

 

[Wolfgang Schechinger]

Ross,

in that case, a genetic analysis might provide more conclusive results (provided that reference data already exists).

Wo
-------- Original-Nachricht --------
> Datum: Mon, 20 Dec 2010 08:14:12 +0000
> Von: Ross Ahmed <[hidden email]>
> An: [hidden email]
> Betreff: Re: Assessing bird colouration

> Many thanks for all the comments so far.
>
>  
>
> I'll just quickly summarise the background to my query. In late November,
> a
> Hoopoe was found dead in Durham, UK and is now stored at the Discovery
> Museum in Newcastle, UK. To my eyes, the dark colouration of the orange on
> the bird suggests it may belong to the subspecies saturata (which
> originates
> from Asia), as opposed to the subspecies epops (which originates from
> Europe). As concluded by Ericson, a key difference between these
> subspecies
> is the colouration of the orange. A photo of the bird can be found here:
> http://www.flickr.com/photos/31121749@N06/5276824748/
>
>  
>
> If it is possible to produce equivalent results as Ericson, this would
> help
> in deciding which subspecies the bird belongs to (and therefore whether it
> may have originated from Asia or Europe).
>
>  
>
> So could anybody provide a step-by-step method, which would produce
> results
> comparable with those of Ericson, for photographing the bird (my camera is
> a
> Canon 40D) and analysing the photos in Image J (or any other program)?
> Robert Dougherty in particular hinted that this may be possible.
>
>  
>
> Thanks again
>
>  
>
> Ross
>
>  
>
> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
> Wolfgang Schechinger
> Sent: 20 December 2010 06:51
> To: [hidden email]
> Subject: Re: Assessing bird colouration
>
>  
>
> If you can *see* the differences between the birds with your eyes in
> pictures taken with a camera that takes RGB images, I think, it should be
> possible. Maybe you need some sort of "standard curve" / reference data
> sets
> created by visual assertion of birds to groups.
>
> Actually this might be a good job for a neuronal network.
>
>  
>
> Have fun
>
>  
>
> Wo
>
> --
>
> GRATIS! Movie-FLAT mit über 300 Videos.
>
> Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome
>
>  

--
GRATIS! Movie-FLAT mit über 300 Videos.
Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome

[Gabriel Landini]

Hi Bob thanks for the follow up.
My argument was that one cannot, with only 3 sensors, for example,
differentiate whether something is truly orange or the sum of certain green
and red (and other) wavelengths. So it is possible that the plumage across
species is in reality very different, but the colour perceived might be the
same. One wants to avoid this: http://en.wikipedia.org/wiki/Metamerism_(color)
And it might be very tricky if the data is gathered as RGB in the first place.

Of course, I agree that to differentiate a green parrot [*] from a white swan,
one might not need spectral analysis in the first place. My suggestion is that
if there is specific equipment to do this correctly, I see no point in trying
to do it otherwise.

Liaising with a physics, optics or engineering department might give the
opportunity to use a spectrometer or a tunable filter and resolve this in the
best possible way. They might be keen in collaborating on a project like this.

But in any case, to further address the photo posted, one needs a calibrated
image shot with a colour chart and use something like this:

http://imagejdocu.tudor.lu/doku.php?id=plugin:color:chart_white_balance:start

I hope it helps.

Cheers

Gabriel

(* either dead or not, :-) in case there are any Monty Python fans)

[Michael Schmid]

Hi,

just one remark that seems to have been omitted in the discussion so  
far:

When using a digital camera instead of spectra, you must make sure  
that the images are taken under equal illumination conditions.

Adjusting white balance will be not enough when using different light  
sources such as sunlight, light bulbs or fluorescent lamps (the  
latter is probably the worst).
The problem of color metamerism mentioned by Gabriel not only implies  
that different spectral distributions are perceived as the same color  
(this might be tolerated if there are only a few different pigments  
or is interference is responsible for the colors).
Even more mimportant, the same actual color will give different R:B:G  
ratios when the spectral distribution of the lamp is changed.


Michael
________________________________________________________________

On 20 Dec 2010, at 09:14, Ross Ahmed wrote:

> Many thanks for all the comments so far.
>
>
>
> I'll just quickly summarise the background to my query. In late  
> November, a
> Hoopoe was found dead in Durham, UK and is now stored at the Discovery
> Museum in Newcastle, UK. To my eyes, the dark colouration of the  
> orange on
> the bird suggests it may belong to the subspecies saturata (which  
> originates
> from Asia), as opposed to the subspecies epops (which originates from
> Europe). As concluded by Ericson, a key difference between these  
> subspecies
> is the colouration of the orange. A photo of the bird can be found  
> here:
> http://www.flickr.com/photos/31121749@N06/5276824748/
>
>
>
> If it is possible to produce equivalent results as Ericson, this  
> would help
> in deciding which subspecies the bird belongs to (and therefore  
> whether it
> may have originated from Asia or Europe).
>
>
>
> So could anybody provide a step-by-step method, which would produce  
> results
> comparable with those of Ericson, for photographing the bird (my  
> camera is a
> Canon 40D) and analysing the photos in Image J (or any other program)?
> Robert Dougherty in particular hinted that this may be possible.
>
>
>
> Thanks again
>
>
>
> Ross
>

[dscho]

Hi,

On Mon, 20 Dec 2010, Gabriel Landini wrote:

> Of course, I agree that to differentiate a green parrot [*] from a white
> swan, [...]
>
> (* either dead or not, :-) in case there are any Monty Python fans)

I thought it was a blue Norwegian parrot...

Ciao,
Dscho

[Gabriel Landini]

On Monday 20 December 2010 15:49:47 Johannes Schindelin wrote:
> Hi,
>
> On Mon, 20 Dec 2010, Gabriel Landini wrote:
> > Of course, I agree that to differentiate a green parrot [*] from a white
> > swan, [...]
> >
> > (* either dead or not, :-) in case there are any Monty Python fans)
>
> I thought it was a blue Norwegian parrot...

No it isn't!

http://www.youtube.com/watch?v=teMlv3ripSM

Cheers and happy holidays,

Gabriel

[Robert Dougherty]

On Dec 20, 2010, at 12:14 AM, Ross Ahmed wrote:
>
> So could anybody provide a step-by-step method, which would produce results
> comparable with those of Ericson, for photographing the bird (my camera is a
> Canon 40D) and analysing the photos in Image J (or any other program)?
> Robert Dougherty in particular hinted that this may be possible.
>
>

Ross,

Here are some possible suggestions based on the discussion so far.  

1. Put your hands on a spectrometer and a tungsten-halogen light, etc., and duplicate  Ericson's approach.  Measure a spectrogram and calculate the CIELab color parameters a, b, and L. Compare these with Ericson's tables.

2. Look at the link Gabriel sent: http://imagejdocu.tudor.lu/doku.php?id=plugin:color:chart_white_balance:start. Get one of the color charts and photograph the bird and the chart at the same time with your Canon. (Same illumination, same camera settings, maybe in the same exposure.)  Use the Yves Vander Haeghen's macro to adjust the image to be "closer to the sRGB color space." Convert to CIELab using Fiji using the Image>Color menu that Johannes mentioned.  Compare with Ericson's tables as in method 1.

Method 1 is probably more accurate, as Gabriel has pointed out, because the spectrometer is a scientific instrument and the camera is not.  In either case, I'd suggest some validation testing. Perhaps measure hoopoes of known origin?

Bob

Robert Dougherty, Ph.D.
President, OptiNav, Inc.
1414 127th Place NE #106
Bellevue, WA 98005
[phone]
FAX [phone]
www.optinav.com
[hidden email]