Do it yourself green on a brown spot

The green-on-brown weed finder design is now available for anyone looking to create their own site-specific weed management tool.

The OpenWeedLocator (OWL) is an open-source, cost-effective image-based approach to the detection of fallow herbs.

Guy Coleman from the University of Sydney, Australia, is working on the project. He said the OWL is an opportunity to redefine the approach to weed detection by enabling community-focused technology development and implementation.

“It’s a box with a camera, a Raspberry Pi, and a set of relay control boards. Since we’ve all published open source information about how it works, anyone can contribute and develop and build one of them, ”said Coleman.

He said the OWL is an improvement on weed detection systems that use reflection-based sensors, as that system uses cameras and image analysis to locate weeds.

“Using cheaper cameras and control boards really shifts things to detection and image analysis to get actionable output that lets you plug a spray solenoid into one of the relay boards or use it to control a targeted tillage tool.”

For example, Australia uses a product called Weed Chipper, which uses specially designed, fast-reacting tines that fall down and work on individual weeds.

“If you want to build a point sprayer, all you have to do is set up the nozzle control so that each nozzle can be switched on or off independently.”

Everything in the modular OWL box requires a 12 volt power source.

The OWL was not developed for connection to the nozzle control systems of existing syringes. Instead, it is the control system for homemade selective weed management tools.

So far, the OWL has been working at speeds of up to eight kilometers per hour.

Coleman said the OWL can be used with syringes mounted on the back of trucks, on a UTV, and research is being carried out on an autonomous platform using the OWL.

The system is installed on an Agerris Digital Farmhand robot for two meter wide spot spraying at the Plant Breeding Institute of the University of Sydney in Narrabri, Australia.

The OWL can be used with any small robotic platform to do the detection and application side of homemade site-specific weed management tools.

“We took the open source path because we are building that end-user experience into development,” said Coleman.

“Another main reason for us was to develop a community around weed detection and site-specific weed control. That would then not only shift where we can see fallow areas, i.e. green on brown. But also in more complex scenarios like in-crop applications, since everything is image-based, ”said Coleman.

He said an inspiration for OWL’s open source development was Brian Tishler, who farms near Mannville, Alta and is the driving force behind Ag Open GPS, an open source software for mapping, guidance and technology Section Control.

“He (Tishler) built a community with people all over the world contributing to the software and designs, telling their stories about what they built and how they took the original ideas and slightly modified them to make them and your contribution can help everyone else, ”said Coleman

Tishler said the green on brown patch spray platform has limited uses in Canada because few people store chemical fallow on the prairies, but there are many growing areas around the world that will benefit from this system.

“The next step in what Guy (Coleman) and of course others are working on, including the people in the Ag Open GPS community. You do a lot with computer vision, that’s also the direction I’m going, ”said Tishler.

“We don’t have to identify the weeds. We just need to identify the crop and remove everything else. I think there is huge potential there. “

There are several companies that are using computer vision and artificial intelligence based systems to identify weeds in real time so that weeds can be identified and managed without operator intervention.

This is tedious work that requires training algorithms to detect weeds and plants by taking thousands of images of weeds and then categorizing the weeds and even the pixels in the images that contain the weeds.

Most weed discrimination algorithms are developed by companies and are proprietary.

However, both Ag Open GPS and OWL seek user participation to identify and categorize weeds, and the resulting algorithms will be available to everyone and will be improved.

The computing power and control systems for individual weed management are already in place. The biggest obstacle to these systems is an archive of weed and plant images that device manufacturers or developers can access.

Carpenter said there was room for the government, grower groups or universities to band together to help build an open weed identification database that developers could use to create site-specific weed management tools.

“Canada could be an absolute star when it comes to putting together a repository and finding a home to collect that data,” said Tishler.

“Could you imagine if the University of Saskatchewan started this project by saying, ‘Okay, we’re following the Ag Open GPS model, where we involve farmers around the world to send in these images.’ What a feather in a hat for any country or province that does that. “