Measuring Color Underwater: Techniques and Challenges

[Fpmagnani]

Hello, I want to measure color of some objects underwater. I believe that one way of do that would be using the RGB system on photos. The other way would be measuring absorbance directly with an spectrometer, but I don't know how I can standardize my conditions. I have to measure the color of these objects but I can't move then to a place with controlled conditions.
Can you guys give me some help in how can I do these measuring (with photos and spectrometer)?

 

[K^2]

Maybe you can describe the setup a little bit better. I'm not exactly clear on what you are doing and why.

If possible, spectral analysis is a better option. RGB system does not represent all possible visually-distinct colors. But again, not knowing why you need this, it's hard to tell whether it'd get in the way.

Finally, water does absorb light differently at different wavelengths, so it will have an effect. Are you trying to correct for it? Or is that the thing you are trying to study?

 

[Fpmagnani]

Yes, I should have explained better. I want to measure color in corals. They are not in the sea, these corals are in farming conditions. In aquariums with a 25 cm water column and under three distinct illumination systens: HQI, fluorescent T5 and solar light. I can't move the corals because, if I touch them, they will certainly shrink and exchange color.

 

[Fpmagnani]

The growth and the color that the corals get is different in each treatment. I want to measure these variation between the treatments. The growth, I already know how to measure, the problem here is the color.

 

[K^2]

Hm. Whether RGB is good enough or if you need spectrum is still going to depend on how you will use your measurements afterwards. Like, if you are looking for a specific, known pigment presence, RGB can probably provide sufficient information. But if you aren't sure what you are looking for yet, you probably need full spectrum.

Good news is that it sounds like you'll be able to calibrate it pretty easily. Just get something very white, submerge it next to some corals (I presume that can be done without causing any damage) and get the spectrum with each type of lighting condition. This way, you can divide the spectrum you get by the calibration spectrum and call it your "actual" spectrum. If RGB is good enough, you'd do the same thing with 3 components.

 

[Fpmagnani]

Great! So, i think that RGB will do that! Thanks!
Anyway, i have an USB4000 spectrometer from ocean optics that i will use to map the irradiance from these lights in the aquariums. It can be used to measure absorbance too, do you think that it is too hard to measure in the conditions presented here?

 

[K^2]

Shouldn't be. You are really dealing with just three factors. How much light was emitted at each wavelength, how much of that was absorbed by water, and how much was absorbed/reflected by corals. So if you get good measurements of the first two, you should be able to figure out the last one easily enough.

 

[Fpmagnani]

The problem is that the corals, depending of the kind of light that they are exposed will produce more or less proteins to reflect the light and can use too, other strategies to absorb or reflect more. And these kind of behavior is what change the color of them, and is what I want to measure

 

[K^2]

Without knowing specifically what proteins they are, and what their optical properties are, I'm not sure RGB will give you enough information.

 

[sophiecentaur]

If you want a good estimate of colour under all sorts of different lighting conditions then I think you should take a reference colour chart down there with you. A grey scale is used by all serious photographers to help with getting the right colour balance but your accuracy would be even better if you compare RGB values of your test colour with those you get from your standard reference colours.
Of course, this will only give the RGB values from a wideband three filter analysis. If you want more detailed information then you need a spectrometer. That would be soooo much more trouble that you should consider the digital camera / RGB / reference chart solution first.

 

[Fpmagnani]

Well, sorry for the late response. I will try to buy one of these charts, but one that be waterproof. Thanks for all the help. When i have news i will post in here!