My project is to detect a specific weed from a corn field. A robot will move inside a corn field, taking video with a Raspberry Pi camera. I have the weed image as "positive" and corn field image as "negative". What are the steps? How do I train a classifier and detect this weed from video?