AI plant disease scanner: local treatment advice, not generic chemistry
Gardener photographs a leaf, AI identifies the disease or pest and suggests a specific UK garden centre product or organic alternative.
You run a mobile app for hobby gardeners and small growers. User photographs an affected leaf, fruit or root, computer vision identifies fungus, virus or pest and suggests a concrete remedy: which products work, where to buy them in UK garden centres, how to dilute and when to apply, plus an organic variant for chemistry-averse users. Freemium model with premium subscription for unlimited photos, plant history and seasonal alerts.
A British gardener grows tomatoes all year, in late summer notices brown patches on the leaves and starts googling. They find conflicting advice on multiple forums, none of which mentions UK regional conditions or local garden centre availability. By the time they figure out whether it is late blight or septoria, the crop is gone.
๐ธComputer vision became commodity in recent years
OpenAI Vision API, Google Gemini Vision and open-source models like YOLOv8 handle plant diagnostics with reasonable accuracy on common diseases. A few years ago this needed a full ML team. Now prompt engineering plus a curated local dataset are enough to build the app at a fraction of the old cost.
