Keeping an AI eye on the birds

The study used a small drone equipped with a camera and computer vision – a form of artificial intelligence – that enables recognition and processing of visual information – to automatically recognise what the turkeys are doing. The research was the first to test whether a drone, boosted with a computer vision model, could automatically detect different turkey behaviours from an overhead video.

The researchers used a commercially available drone with a regular colour camera to record 160 young turkeys, aged between 5-32 days, 4 times a day at the Penn State Poultry Education and Research Centre. The trajectory of the drone was designed to ensure full area coverage from the camera footage during each flight.

Dataset of behaviours

Researchers took individual image frames and manually labelled the behaviour of the birds. They then created a dataset of more than 19,000 examples of specific behaviours, including:

• Feeding
• Drinking
• Sitting
• Standing
• Perching
• Huddling
• Wing flapping

The team then used the images to train, test and validate a computer vision model called YOLO – you only look once – commonly used to detect objects and actions in images.

Detecting specific behaviours

The researchers tested several YOLO versions and found that the best model could correctly find 87% of all present behaviours and accurately detect specific behaviour 98% of the time.

Senior author Enrico Casella, assistant professor of data science for animal systems at Penn State, said that these metrics are good, especially for behaviour classification in a real farm environment, which often is visually messy and challenging.

“This work provides proof of concept that drones plus AI can potentially become an effective, low-labour method for monitoring turkey welfare in commercial production,” Casella said. “It lays the groundwork for more advanced, scalable systems in the future.

The study shows that a drone-equipped AI system can accurately detect turkey behaviours, he added. “This method could reduce labour demands; it could allow continuous, non-invasive monitoring of bird welfare in commercial farms and it may also reduce the need for constant human presence, lowering training and staffing burdens.”