Hello everyone! I am on my final year and we're trying to create an automated device that can trace green color or "algae" on the surface of the water. May I ask for suggestions on how can we trace the algae or the green color on the surface of the water as our device aims to collect them by following them on the surface, if possible, through sensors? I am just a beginner but I will surely try my best to understand the concept and accomplish this project as this will determine if I will finish my studies on time. Thank you so much for your help and have a nice day!
if you were to take a picture of the surface of the water, could you by looking at the picture tell if there is algae ?
You could try using an RGB color sensor, like the TCS34275:
To reliably detect algae (cyanobacteria) might require higher color resolution, like this spectral sensor which resolves 6 bands:
Since you will be detecting reflected light, the results will depend VERY strongly on the illumination. Reflected illumination from clouds versus sunlight will give different spectra.
May I propose you another approach?
the properties of the water surface and the water surface with algae may have different properties you can use, without the need of a color camera (and so avoiding any issue with lightning the scene or ambient light.
For example, a sound wave may be partially absorbed by the water but reflected by algae (or any junk floatting near-by). You lost the ability to discriminate the color (but I feel you didn't need to determine for brown or red algae, just population at the surface) but you can imagine a simpler apparatus.
I have in mind 2 differents configurations (including easy stuff like speaker/micro or ultrasonic device):
Fisrt the emitter and receiver are side by side, suspended above the surface (it will fit with the ultrasonic sensor)
Second, the emitter and receiver are opposed in a way the emitted wave will make an anlge of 45° with the surface. It will fit if you have distinct devices for emitting and detecting.
Hope you will make a great project!
EDIT: actually a light beam (no specific color required) directed at 45° to the surface could give you better results than sound as algae WILL reflect it and water may mainly absorb it.
Correct.
A polarizing filter may be required.
The polarization of the sky
The spectrum of the illumination matters much more than the polarization.
Which leads one to envision 2 separate sensors, one to view the sky and one to view the surface reflection of the water. My suspiction is that the water detector would need a mast to elevate it over the water with sufficient height that floating algae could be spectrum detected.
re: Absorption / reflection of sunlight - Understanding Global Change (berkeley.edu)
Interesting paper on the illumination spectrum of overcast sky: it is not often as "gray" as most people seem to think: https://www.researchgate.net/publication/7564666_Colors_of_the_daytime_overcast_sky
My initial thought was a a pipette on a servo with a color sensor looking inside the pipette which could be shielded from natural light, reflection and general weather disturbance
If the picture is representative of what you would get, extracting the Dominating colors and building areas could work. May be a better task for a RPi