An international team of scientists from the University of Bath, the University of Oxford and the University of Twente in the Netherlands have successfully used satellite cameras coupled with deep learning algorithms to track the movements of African elephants.
The population of African elephants has plummeted over the last few decades thanks to poaching and loos of habitat. The species is now classified as endangered with just 50,000 left in the wild.
Currently, conservationists monitor the populations of endangered and under threat animals such as elephants by counting them one-by-one from low-flying aeroplanes.
In this study the team used an automated artificial intelligence system created by Dr Olga Isupova, a computer scientist at the University of Bath, to analyse high-resolution images of the elephants as they moved through forests and grasslands captured by the commercially run Worldview-3 observation satellite. They found that their system was able to pick out the animals with the same accuracy as human analysts.
Read more about wildlife:
- Can we save the insects?
- Zoology in 30 seconds: conservation and extinction
- Northern white rhino embryo could save all-female subspecies from extinction
Though the combination of satellite imagery and deep learning has been previously used to identify marine animals, the study marks the first time the technique has been used to monitor animals moving through a diverse, heterogeneous landscape that includes areas of open grassland, woodland and scrub.
“This type of work has been done before with whales, but of course the ocean is all blue, so counting is a lot less challenging. As you can imagine, a heterogeneous landscape makes it much hard to identify animals,” said Dr Isupova.
“Accurate monitoring is essential if we’re to save the species. We need to know where the animals are and how many there are.”
The team chose to run their pilot study using African elephants as they are the largest land animals and therefore the easiest to spot. However, the researchers are hopeful that the technology will be successful in observing other species from space in the future.
“Satellite imagery resolution increases every couple of years, and with every increase we will be able to see smaller things in greater detail. Other researchers have managed to detect black albatross nests against snow," said Dr Isupova.
"No doubt the contrast of black and white made it easier, but that doesn’t change the fact that an albatross nest is one-eleventh the size of an elephant. We need to find new state-of-the-art systems to help researchers gather the data they need to save species under threat.”
Reader Q&A: Do elephants really never forget?
Asked by: Theo Hunter (Aged 11), Sheffield
An elephant has a very large brain for its size and the ‘temporal lobe’ region responsible for memory is more developed with a greater number of folds – this results in powerful abilities to ‘download’ important survival data such as where to find food and water, and who is friend or foe.
The matriarch of a herd (who can live for 60 years) may recognise over 200 individual elephants and can react to the call of a deceased member of her herd two years after their death. During droughts, these grandma elephants lead family members to waterholes by recalling detailed maps they’ve made spanning hundreds of kilometres. So although they undoubtedly forget what they don’t need to remember, they appear to remember what they cannot afford to forget!
Read more about elephants: