What I want to do
Hi all, I'm following your site form my project. I want to use UAVs for crop monitoring with cheap tools available for everyone. I found this paper:
It looks really promising, since it shows that what I had in mind was possible. It shows how to combinate the 3 camera channels to get the best NDVI possible, in particular on a Canon 500D/350D, recreating the real NIR and R channels as a linear combination of the R G B channels. Without a sort of calibration about the sensitivity of each channel, the NDVI won't be really scientifically reliable, since each channel will get an amount of light and translate it into brightness according to its sensitivity, which will be unknown.
The paper shows how to overcome that problem. I really would love to recreate the procedure with a more light and cheap camera, like a normal Canon powershot or something like that, but I don't have the necessary gears to follow the whole procedure.
Questions and next steps
Any of you knows if the equation showed in the paper could be useful also with other compact Canon model without any modification. I wonder if it could be possible, since the sensor between the 350D and the 500D is different, so maybe it is a matter of sensor producer.