From above, the green pixels will obscure the brown ones, so the best you could say is "there are 3 forests in this area". If you are under the canopy itself, y

From above, the green pixels will obscure the brown ones, so the best you could say is "there are 3 forests in this area". If you are under the canopy itself, you have much larger problems. The colours are very similar, so distinguishing between one tree trunk and another is not the same as distinguishing between a white licence plate and a blue car. The shapes are very similar, so distinguishing between one tree and another is not the same as distinguishing between a car and a house. Some trees will be obscured by others so you will need a number of different cameras to create a 3D map. You would also need to measure the distance between the cameras to determine what area they cover. A 10 minute analysis revealing a list of complexities as long as my arm is not a good start. It will probably be easier (let alone more accurate) to just count the trees.