I am interested in learning more about the capabilities of AI, one of my ideas with practical functionality is using images of the rear of a log hauling truck to measure the individual logs using AI. The diameter and length of the log determine the board foot of that log. The diameter is the variable, with most lengths being the same 8ft or 16ft length. The AI would have to measure the diameter usually in 1-inch increments. I assume you would have to train the AI using pictures of logs that were manually measured. The images would mostly be straight on rear shots, not much off angle. However not every tree grows perfectly round so it will need to be able to measure those to maximize board feet.
Using AI long term it would be interesting if it could suggest the best way to cut the log gaining the most lumber board feet from the log.
What would be used to develop something like this? Obviously some type of image recognition with learning.