Aarhus Universitets segl

Drone spots game in the field

Drone spots game

in the field

The number of animals killed by machinery on agricultural land is going to be reduced with the use of drones, cameras and algorithms. Researchers at Aarhus University are behind a project that has great potential for both animal and farmer.

very year thousands of deer, hares and pheasants die in the fields when they get run over by large farm machinery at harvest time. In particular, the accidents occur during the summer when the animals migrate to the edge of wooded areas. But according to the first results from an Aarhus University project to monitor deer, the problem can be greatly reduced by using drones equipped with thermal cameras.

Advanced algorithms are lifesavers
In partnership with two companies, researchers have conducted the first flights over agricultural land using a drone equipped with a thermal camera. In one of the flyovers, the drone managed to find two roe deer fawns despite only covering a few hectares of land.

The prospects for using drones to monitor game are substantial, explains the project’s anchorman, Senior Researcher Rasmus Nyholm Jørgensen:

"In the future we will be able to provide a solution for the farmer where wildlife in the field will be pinpointed with great accuracy. The problem today is that the animal will usually lie flat on the ground when an agricultural machine is approaching. Therefore it is extremely difficult to detect the animal and prevent a collision. Even for a hunter with a trained dog, it is an almost impossible task."

He also points to another problem related to wildlife killed in the field, which is that of food quality. The game may be infected with disease-carrying bacteria and contamination of the harvest can lead to bacterial poisoning for whoever ends up eating it.

Bacteria in forage grasses involve a risk of infecting entire cattle herds.

Stressful accidents
Another negative consequence of the many collisions is that the experience is often very unpleasant for the driver of the machine.

"The animals are hit by large agricultural machines that may be driven by an 18-year-old young man who subsequently has to go out to deal with the animal and maybe finish it off. It is bound to affect you psychologically," says PhD student Kim Arild Steen.

Researchers at Aarhus University have recently been looking for solutions to minimise the number of animals run over by agricultural machinery in the field. With the latest use of drones, there is now a good chance of finding an optimal solution that can alleviate the problem of so many animals being killed.

"The advantage of using drones is that an aerial scanning can be very fast. We can already monitor large areas in a very short time," explains Rasmus Nyholm Jørgensen.

Cameras detect body heat
Although researchers have made great strides towards finding a suitable solution, Rasmus Nyholm Jørgensen recognises that there is still a long way to go before they have a market-ready product.

"We have developed algorithms that are capable of detecting the presence of animals in the photos from the thermal camera. However at this relatively early stage of development, we still have problems in distinguishing between objects in the field that have been warmed by the sun and animals that emit heat. Our ambition is to develop a system that is able to distinguish between animals and other objects," he explains.

Rasmus Nyholm Jørgensen hopes that the project will help lead to a workable solution, ultimately reducing the number of animals killed on agricultural land.

"Many farmers experience these fatal situations and would really like to solve the problem. And we would really like to help them and believe that we can do so," says Rasmus Nyholm Jørgensen.

How the drone spots animals in the field

The drone is equipped with a heat-sensitive camera. From the air it is able to detect body heat, for example from game.

The researchers have developed the algorithms that can detect the presence of animals via the camera readings.

Now the challenge is to develop the algorithms that will allow the drone to tell the difference between objects in the field that have been warmed by the sun and animals lying in the grass.