3D Orchards: UC Researcher Turns Farms Into Virtual Reality
Farmers are relying more and more on technology to help them manage their crops and often that means working with unmanned aerial systems. Using drones to make two dimensional maps of orchards isn’t anything new, but one agricultural researcher – Ali Pourreza – in Central California is taking existing drone technology to the next level.
"It's a virtual reality actually. So basically you can put on one of those goggles and walk through a virtual orchard." - Ali Pourreza
“I thought, okay, two-dimensional imaging has been around a long time and it's helped a lot, but right now we have the capability to make 3D models,” says Pourreza.
In Pourreza’s hands is a remote control hooked up to his iPhone. He’s flying a little white drone over a field of nectarines near Parlier, south of Fresno. The camera on the drone is collecting images as it flies back and forth over the field.
“So in real time you can see what’s going on,” says Pourreza, an agricultural engineer at the Kearney Agricultural Research and Extension Center. “It goes up to 200 feet. That’s the target altitude. Then it goes to the starting point.”
The images and data he’s collecting will later run through a computer system that creates a 3D map of the field. It takes those 2D images and puts them in 3D.
“It’s a virtual reality actually,” Pourreza says. “So basically you can put on one of those goggles and walk through a virtual orchard. So this is something interesting even for people who are not in agriculture.”
He says he got the idea earlier this year at a conference when he realized no one else was utilizing drones in this way on farms.
“I saw that they use this technology to reconstruct an accident and I was impressed that no one had used this for agriculture,” Pourreza says. “I thought it has potential, but I didn’t know until I created my first 3D map.”
The 3D maps Pourreza is creating are almost like a video game. With this technology a farmer can sit at his kitchen table, but also be virtually walking through a field and zooming in to parts of the orchard that look stressed. A farmer can even look at a time lapse of that exact spot and then make a decision on things like irrigation and pesticide use. Pourreza says these virtual vineyards hold a lot of promise, because they cut down the time a farmer or crews need to spend in the field.
“I think virtual orchards opens a lot of doors to collect more precise information and convert them to valuable knowledge that will result in good decisions for growers,” Pourreza says.
But there is an issue, it can cost around $100,000 to use existing remote sensing technology – like LIDAR – to do this. Pourreza says he’s developing a low-cost alternative with a drone equipped with a camera, a computer program to process the data and a smartphone. Hanford Pistachio Farmer James Nichols is testing out the virtual orchard technology.
“There’s always going to be a need to go out into the field, but I think this is going to be able to quantify things that we have never been able to do in the past,” says Nichols. “You’d be able to measure how much growth you have in an orchard from year to year.”
Nichols says the computer software brings all the images together creating a 3D map with all sorts of layers. For instance, with a click the trees can be removed and just their shadows are left. The more light shining through the canopy, the more water the orchard needs.
“We don’t know how much water younger trees use,” says Nichols. “And from that we’d be able to better dial in how much water to give our pistachio and almond trees.”
A 160-acre field can take just under two hours to map. Dr. YangQuan Chen runs The MESA lab focused on drone research at UC Merced. He says this technology once fully fleshed out may greatly alter how farmers manage their crops.
“I think [Ali Pourreza] is one of the earliest proponents for this technique,” Chen says. “It’s a very early stage and I believe it is quite pioneering for precision agriculture.”
But at the same time, Chen says the virtual mapping technology needs a lot of testing before farmers are going to trust the data it computes.
“More research is needed to make sense of the data,” Chen says. “The question probably is the cost. Are we getting less and paying more or paying less and getting more? That’s not very clear.”
Back in Parlier, the drone is about finished collecting data. Ali Pourreza will take the thousands of images back to his lab where they’ll be stitched together into a 3D model of the orchard. In his office he’s showing me an animation of an orchard.
“So you can now see that we have the information from all sides of the tree. As you move your position, which is the camera position, you can see how the tree looks. How tall it is, how wide it is.”
Pourreza says this is just the beginning of using virtual orchards to help farmers make clearer decisions about their crops. He says he’s willing to work with all sorts of researchers to see how the data he collects can help not just farmers, but irrigation specialists and soil scientists.