I need to train a deeplearning model, I need to make the model figure out different plants, like maize and wheat, but they look like these: maize example wheat example
I think it's hard to make a deeplearning model like darknet(now I'm using yolov4), does anyone have some good idea? I don't know how to finish this task.
If detect is useful, which annotation is correct? whole img or split every single feature ?
I used yolov4-tiny, yolov4, but they all perform not good, I have no idea about this