Weed or crop image classifier
Inspired by publicly reported work at John Deere
Tell weeds from crops in field photos so a sprayer could treat only the weeds.
The brief
John Deere has described See & Spray, which uses cameras and machine learning on the sprayer boom to tell weeds from crops and spray only the weeds. Build the image classifier part: train a model on public weed images, measure it per species, and work out how much herbicide spot spraying could save. Good looks like: per-species accuracy, a confusion matrix of the weeds it mixes up, and a simple savings estimate with its assumptions.
Read the case first
What John Deere has said in public about this kind of work, on its own pages:
Make it your own
Do it for a crop near you: paddy, cotton or chilli, with weeds common in your district. Photograph 200 plants yourself (with the farmer's permission) and test the model on them.
AITeky has no link with John Deere, which has not reviewed or endorsed this brief. The brief only draws on what the company has published, uses open data rather than its data, and is a learning exercise: don't present your project as John Deere's work or system.
What you will build
- Explore weed species and image conditions
- Fine-tune a small CNN
- Evaluate per species and analyse mix-ups
- Estimate herbicide saved by spot spraying
- Build a demo that labels a photo
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.