Python Forum

Hello,

I have few UAV images and I am searching how to do 2 things.
1) Identify specific objects on ground and isolate these objects as new images
2) Use the new images from the step 1 in order to train a CNN and output results.

I believe that I should "tell" the CNN which images are what, in other words to put labels on them (is there any better/faster way?) The problem is step 1. How can I achieve this object detection?

this paper may be of interest, I think it's pretty close to what you're looking for:
https://scholar.google.com/scholar?hl=en...ython&btnG=

googling this: Object detection on ground via the use of 'geo/sat UAV images' python
bring us several interesting articles from banana plants to swimming pools. Didn't see people, but there might be something there.

Thanks!