The best tools come from the people who use them.
Noktura is an open ecosystem for vision AI in agriculture. We bring together open hardware, community data, and shared knowledge so anyone can build weed detection that works for the conditions they face.
What Noktura does
Upload your field photos with the agronomy that matters. What crop, what growth stage, what conditions. Keep them private, share with your group, or contribute to public datasets. Detection that works on your land starts with the data you put in.
Image storage
Field photos with the agronomy attached: crop, growth stage, location, weather, and soil. All searchable later.
OWL device management
Register your OpenWeedLocator devices and track what each one is running.
Group sharing
Share datasets with your research group or organisation. Control who sees the GPS coordinates. Publish whenever you are ready.
Public datasets
Browse and download what the community has shared. Filter by region, crop, or growth stage.
Model sharing
Share weed detection models, or find ones already trained for your weeds and crops.
Model trainingComing soon
Train models on community data. From simple green detection through to deep learning.
Agriculture is full of niches
Thistles in a Danish barley field are a completely different challenge to ryegrass in an Australian wheat paddock. The weeds look different, the soils are different, the light is different. A detection model built for one rarely works well in the other.
Farmers have always been innovators, adapting tools and techniques to the specific conditions on their land. But most agricultural technology still comes as a black box, unable to be adapted to fit these niches.
Noktura takes a different approach. Instead of waiting for a corporate solution that might not fit, we give people the tools to build and share their own. Open hardware you can modify. Data you can contribute to and download from. And a community that benefits every time someone shares what works in their corner of the world.
What the community is sharing
Public datasets from farmers and researchers around the world.
ATV-OWL - Barley #2
Images collected with an OWL camera (Rasperry Pi IMX296 Global Shutter) over a later tillering stage barley crop in canola stubble. The OWL was mounted on an ATV on a farm near Manangatang in Victoria, Australia. Images collected by Brad Plant.
ATV-OWL - Lentils #1
Images collected with an OWL camera (Rasperry Pi IMX296 Global Shutter) over a lentil crop at an early growth stage on wheat stubble. The OWL was mounted on an ATV on a farm near Manangatang in Victoria, Australia. Images collected by Brad Plant.
ATV-OWL - Peas on Wheat Stubble
Images collected with an OWL camera (Rasperry Pi IMX296 Global Shutter) over a pea crop at the early stage on heavy wheat stubble (stripper front harvested). The OWL was mounted on an ATV on a farm near Manangatang in Victoria, Australia. Images collected by Brad Plant.